|
|
Post by marcl on Feb 4, 2021 6:40:56 GMT -5
First, it's best to use the miniDSP for all the subs because it allows you to do time and level alignment in the miniDSP, so that Dirac just has to align the combined virtual sub to the other speakers. As for the question of using Center Sub LFE or not, my opinion is if you are going to send small speaker bass to subs ... then use Center Sub Mono to miniDSP Input 1, align all four subs together in the miniDSP, and place them in the four corners of your room (or as close as you can). The latter bit of advice is per the Harman paper by Welti on placement of multiple subs. Now, to time align the four subs in the miniDSP you would use the Impulse Response measurement in REW. Sweep the four subs from 30-100Hz and read the timing difference from the first (usually smaller) peak in the Impulse Response. Add delay to the slower subs until they align. Also, adjust the output levels of all four subs until their levels are equal. If you go for Option #2 with the intention of sending small speaker bass to large front speakers, this can work really well because the front speaker bass will likely be a better match for your small speakers . Still connect Center Sub LFE to Input 1 on the miniDSP and all four subs to the four outputs. In order to get front speaker deep bass to the subs below their range (i.e. like <40Hz) you can put a Y cable on your L/R front outputs and then combine one side to miniDSP Input #2. In the input PEQ put a 40Hz LPF, then send that bass to the four subs in addition to the LFE. View AttachmentWith this arrangement you get full range fronts with the benefit of all four subs. And Dirac sees the fronts and full range and ensures smooth transition from the subs to the fronts at the crossover. You will find a lot more discussion and other options at ttocs Finding Subwoofers thread emotivalounge.proboards.com/thread/57730/subwooferHow do you split the front left and right xlr ( balanced) output to rca (unbalanced)? The minidsp input using rca (unbalaced) connection. Oh okay you have the miniDSP 2X4 HD. Mine is 2X4 Balanced. But today you must be using an adapter cable to go from XLR to RCA. So you can get RCA Y cables to do this. I happen to use balanced splitters. A while back - and for a different reason - someone asked Emotiva if it was okay to combine two outputs with a Y cable and they said it was okay. So you could do it that way. Here's a Female XLR to Male RCA www.amazon.com/Y-Cable-Unbalanced-Y-Splitter-Adapter-Duplicator/dp/B083JHPDV1/ref=sr_1_3?dchild=1&keywords=Dual+RCA+to+XLR+Female+Y&qid=1612438796&s=industrial&sr=1-3 |
|
ttocs
Global Moderator 
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,118
|
Post by ttocs on Feb 4, 2021 12:12:33 GMT -5
Dirac and Levels There has been some confusion about how Processor Levels are affected or not by Dirac. So I did a test to point out what "zero" is before and after Dirac. Zero always remains zero. So Dirac does not change Levels, we do. The following shows what happens when Processor Levels are set to 0.0dB and -12dB, before and after Dirac filter is loaded. These are the settings I used for this test.   I set the Left speaker to -12dB and ran Dirac. This is a screenshot of the Volume Calibration screen in Dirac. Please note that while the Left speaker is set at -12dB in the XMC Levels, Dirac doesn't care and completely disregards the XMC setting.  This spreadsheet shows the actual measured output using the settings above. Notice that while Dirac apparently changes the Overall System Output by a tiny amount, it does not change the relationship between the channels as it relates to Levels Settings in the processor so -12dB remains a vastly lower output level and must be raised back to 0.0dB to match the other speaker.  Just to be clear. Dirac does nothing - relatively speaking - to Levels. 0.0dB in Processor Levels Settings will ALWAYS be 0.0dB and is 0.0dB relative to other speakers. |
|
|
|
Post by p4t on Feb 4, 2021 12:15:45 GMT -5
Thank you, for your detail explanation about different way connecting the subwoofer. I will give it a try. Already ordered a pair of KEF KC62 cause is limited space I have, and also want to try how good these subwoofers is. |
|
|
|
Post by Percussionista on Feb 4, 2021 16:50:33 GMT -5
Confused about the Dirac licensing. I've got an XMC-2, Dirac EDNIB and laptop connected up to the same network, supplied mic plugged into the laptop, Dirac software downloaded, and everybody can see each other. I can select the XMC-2 from within the Dirac s/w, select the (downloaded) microphone's profile, and continue further into the program. I haven't run anything, just wanted to make sure things were hooked up and talking.
So... I have an account on dirac.com, and if I log-in and click on the box for "My Licenses", it comes back with "You currently don’t have any valid licenses". If I instead click on the box "Activate License", it prompts me for a "License Key" in order to claim the license. I have no idea what that might be as I haven't heard from anybody regarding same - In searching through this thread I did see mention that the License number is stored somewhere in the Emo box (XMC-2).
So... have I missed one more step? It appears the s/w and h/w all work together, but... ??
|
|
|
|
Post by marcl on Feb 4, 2021 16:53:24 GMT -5
Confused about the Dirac licensing. I've got an XMC-2, Dirac EDNIB and laptop connected up to the same network, supplied mic plugged into the laptop, Dirac software downloaded, and everybody can see each other. I can select the XMC-2 from within the Dirac s/w, select the (downloaded) microphone's profile, and continue further into the program. I haven't run anything, just wanted to make sure things were hooked up and talking.
So... I have an account on dirac.com, and if I log-in and click on the box for "My Licenses", it comes back with "You currently don’t have any valid licenses". If I instead click on the box "Activate License", it prompts me for a "License Key" in order to claim the license. I have no idea what that might be as I haven't heard from anybody regarding same - In searching through this thread I did see mention that the License number is stored somewhere in the Emo box (XMC-2).
So... have I missed one more step? It appears the s/w and h/w all work together, but... ??
There is no license key for the XMC/RMC. It's your Mac address. If you got this far, you are good to go! |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,118
|
Post by ttocs on Feb 4, 2021 17:04:11 GMT -5
+1 ^^^^^^ No license needed, no login needed, the current operational program does not post any info on Dirac's web site for Dirac 3. Other than a repository for being able to download the latest version, and to report bugs and contact Dirac Support, www.dirac.com is irrelevant to being able to use Dirac 3 on our processors. |
|
KeithL
Administrator
Posts: 10,247
|
Post by KeithL on Feb 4, 2021 17:56:28 GMT -5
That's not exactly true...
You don't need to log in or anything like that. So you don't need a user name or password. However your processor still has a license and the Dirac software still checks that when you run it... The whole process is just transparent to the user now.
Every time the software starts, and connects to your processor, it fetches the equivalent of a serial number... (It's a magic number that is derived from the MAC address of your processor.)
The Dirac software then checks with the servers at Dirac to make sure that your processor has a license... And, if not, then the software won't run... (And you'll get a nice cheerful error message...)
The whole process is tied to a number that is derived from the MAC address of your processor. And that never changes (unless your processor is sent in for repair and the Ethernet interface hardware gets changed).
However, the Dirac software still needs to be able to reach the Dirac servers via the Internet to verify your processor's license, or it won't run.
But, no, the guys at Dirac don't have your name... And you don't have an account there... (but your processor does).
+1 ^^^^^^ No license needed, no login needed, the current operational program does not post any info on Dirac's web site for Dirac 3. Other than a repository for being able to download the latest version, and to report bugs and contact Dirac Support, www.dirac.com is irrelevant to being able to use Dirac 3 on our processors. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,118
|
Post by ttocs on Feb 4, 2021 18:06:58 GMT -5
That's not exactly true...
You don't need to log in or anything like that. So you don't need a user name or password. However your processor still has a license and the Dirac software still checks that when you run it... The whole process is just transparent to the user now.
Every time the software starts, and connects to your processor, it fetches the equivalent of a serial number... (It's a magic number that is derived from the MAC address of your processor.)
The Dirac software then checks with the servers at Dirac to make sure that your processor has a license... And, if not, then the software won't run... (And you'll get a nice cheerful error message...)
The whole process is tied to a number that is derived from the MAC address of your processor. And that never changes (unless your processor is sent in for repair and the Ethernet interface hardware gets changed).
However, the Dirac software still needs to be able to reach the Dirac servers via the Internet to verify your processor's license, or it won't run.
But, no, the guys at Dirac don't have your name... And you don't have an account there... (but your processor does).
+1 ^^^^^^ No license needed, no login needed, the current operational program does not post any info on Dirac's web site for Dirac 3. Other than a repository for being able to download the latest version, and to report bugs and contact Dirac Support, www.dirac.com is irrelevant to being able to use Dirac 3 on our processors. So then let me amend my comment. Once I download Dirac 3.x.x, www.dirac.com becomes irrelevant "to me". |
|
klinemj
Emo VIPs 
Official Emofest Scribe
Posts: 15,072
|
Post by klinemj on Feb 4, 2021 18:56:17 GMT -5
Confused about the Dirac licensing. I've got an XMC-2, Dirac EDNIB and laptop connected up to the same network, supplied mic plugged into the laptop, Dirac software downloaded, and everybody can see each other. I can select the XMC-2 from within the Dirac s/w, select the (downloaded) microphone's profile, and continue further into the program. I haven't run anything, just wanted to make sure things were hooked up and talking. So... I have an account on dirac.com, and if I log-in and click on the box for "My Licenses", it comes back with "You currently don’t have any valid licenses". If I instead click on the box "Activate License", it prompts me for a "License Key" in order to claim the license. I have no idea what that might be as I haven't heard from anybody regarding same - In searching through this thread I did see mention that the License number is stored somewhere in the Emo box (XMC-2). So... have I missed one more step? It appears the s/w and h/w all work together, but... ??
Building on marcl's post...have you submitted your MAC address to Emotiva? If you have not - you need to do that. Once they have that and Emotiva submits that to DIRAC and DIRAC does its part, you can run DIRAC. As far as logging into your DIRAC.COM account - last time I checked, I didn't have any valid licenses either - but my DIRAC runs. Net, as long you you have submitted your MAC address and waited a few days for Emotiva and Dirac to activate your new XMC-2 for DIRAC, all should be fine. Mark |
|
|
|
Post by marcl on Feb 5, 2021 12:30:33 GMT -5
I wanted to share some ideas on measurement patterns. With Dirac 1 we could only do 9 positions, but with Dirac 3 was can choose 9, 13, or 17 and within each we can choose to eliminate some. I have a unique (to say the least!) speaker position situation with my fronts and so I have to compensate in some ways so as not to befuddle Dirac. Here's the arrangement that's working well now. I use the 13 position Focused pattern, but eliminate the two low rears and two outer centers, and I modify the other measurements into a V shape.  I do the measurements with the two couch sections around the MLP fully reclined and covered with blankets. For the MLP #1 measurement only, I place two 2x4ft absorbers perpendicular to the near edge of each of the front speakers, and propped a couple feet off the floor. The purpose of this is to let Dirac get the distance and Impulse Response measurement without any stray sound from the front or rear of the dipole Magnepan 3.7's. I remove the absorbers for the remaining measurements. The pattern still considers that the fronts reflect off the side walls, hence the V shape of the pattern. I eliminate the rear low positions due to the height of the couch back. The results of 7.1 measurements ....    The waterfall is left front only, but is indicative of the overall time domain response, especially below 200Hz. The reason why the center and surrounds measure higher in the bass is a bass management issue due to the bass being played by both fronts. Hopefully this will be corrected in a future firmware release. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,118
|
Post by ttocs on Feb 5, 2021 14:05:10 GMT -5
I wanted to share some ideas on measurement patterns. With Dirac 1 we could only do 9 positions, but with Dirac 3 was can choose 9, 13, or 17 and within each we can choose to eliminate some. I have a unique (to say the least!) speaker position situation with my fronts and so I have to compensate in some ways so as not to befuddle Dirac. Here's the arrangement that's working well now. I use the 13 position Focused pattern, but eliminate the two low rears and two outer centers, and I modify the other measurements into a V shape. I do the measurements with the two couch sections around the MLP fully reclined and covered with blankets. For the MLP #1 measurement only, I place two 2x4ft absorbers perpendicular to the near edge of each of the front speakers, and propped a couple feet off the floor. The purpose of this is to let Dirac get the distance and Impulse Response measurement without any stray sound from the front or rear of the dipole Magnepan 3.7's. I remove the absorbers for the remaining measurements. The pattern still considers that the fronts reflect off the side walls, hence the V shape of the pattern. I eliminate the rear low positions due to the height of the couch back. The results of 7.1 measurements .... The waterfall is left front only, but is indicative of the overall time domain response, especially below 200Hz. The reason why the center and surrounds measure higher in the bass is a bass management issue due to the bass being played by both fronts. Hopefully this will be corrected in a future firmware release. That's interesting about the V shape. So having intermediate mic points in the "mouth" of the V caused some issues? Also, must be nice to have such a quiet room! Right now, with fridge and HVAC off, the noise floor is 45dB(C) according to REW's SPL app. |
|
|
|
Post by marcl on Feb 5, 2021 14:13:44 GMT -5
I wanted to share some ideas on measurement patterns. With Dirac 1 we could only do 9 positions, but with Dirac 3 was can choose 9, 13, or 17 and within each we can choose to eliminate some. I have a unique (to say the least!) speaker position situation with my fronts and so I have to compensate in some ways so as not to befuddle Dirac. Here's the arrangement that's working well now. I use the 13 position Focused pattern, but eliminate the two low rears and two outer centers, and I modify the other measurements into a V shape. I do the measurements with the two couch sections around the MLP fully reclined and covered with blankets. For the MLP #1 measurement only, I place two 2x4ft absorbers perpendicular to the near edge of each of the front speakers, and propped a couple feet off the floor. The purpose of this is to let Dirac get the distance and Impulse Response measurement without any stray sound from the front or rear of the dipole Magnepan 3.7's. I remove the absorbers for the remaining measurements. The pattern still considers that the fronts reflect off the side walls, hence the V shape of the pattern. I eliminate the rear low positions due to the height of the couch back. The results of 7.1 measurements .... The waterfall is left front only, but is indicative of the overall time domain response, especially below 200Hz. The reason why the center and surrounds measure higher in the bass is a bass management issue due to the bass being played by both fronts. Hopefully this will be corrected in a future firmware release. That's interesting about the V shape. So having intermediate mic points in the "mouth" of the V caused some issues? Also, must be nice to have such a quiet room! Right now, with fridge and HVAC off, the noise floor is 45dB(C) according to REW's SPL app. About 43dbC at the MLP at the moment but sometimes the refrigerator makes a couple db more noise. I always see it in the Waterfall at 40Hz. With HVAC and refrigerator off I pick up another 3-4db but I don't find it makes any difference to the measurements. Moving the computers and NAS boxes away from the MLP I was able to pick up about 6db. The reason for the tight arrangement and the V is to keep the mic inside the envelope of the shape of the angle of the fronts, not outside. The actual direct sound of the fronts that I listen to is bounced off the side walls, but if I get the mic too far left and right of MLP it also hears some of the direct (not bounced) sound from the side of the speakers facing the walls. This causes Dirac to get extraneous sound 1-2ms out of sync that does not contribute well to the listening at the MLP. When I had the speakers in a conventional front-facing and toed in arrangement, I used a wider rectangular pattern across both cushions on the couch, with good results. |
|
|
|
Post by leonski on Feb 5, 2021 18:06:47 GMT -5
Marci, When doing your initial setup of your Maggies, did you experiemnt with ALL FOUR logical possibilities?
That we be tweeter IN or OUT......a simple L/R swap of the speakers.
But most people don't know that in years past, Magnepan was sold POLE PIECE to the listener, which was changed in the 90s, sometime.
That gives another 2 logical setup possibilities with pole piece TO the listener.
When I set up my 1.6s, I went thru ALL the possibiliies. Some setups lasted 30 seconds. Others a couple hours. Still others took a few days.
I finally settled and than took detailed measurments so I could get 'em back to best spot after vacuuming or whatever.
|
|
|
|
Post by marcl on Feb 5, 2021 18:37:26 GMT -5
Marci, When doing your initial setup of your Maggies, did you experiemnt with ALL FOUR logical possibilities? That we be tweeter IN or OUT......a simple L/R swap of the speakers. But most people don't know that in years past, Magnepan was sold POLE PIECE to the listener, which was changed in the 90s, sometime. That gives another 2 logical setup possibilities with pole piece TO the listener. When I set up my 1.6s, I went thru ALL the possibiliies. Some setups lasted 30 seconds. Others a couple hours. Still others took a few days. I finally settled and than took detailed measurments so I could get 'em back to best spot after vacuuming or whatever. I had Tympani 1D's for 35 years and always had them tweeters-out because I didn't have space to try them the other way. I've had the 3.7's 9 years and have tried them tweeters-in and tweeters-out, though I prefer to call it woofers-in and woofers-out because tweeter placement sets the soundstage based on distance to the MLP ... then you either put the woofer inside or outside. In conventional setup I liked them very close to the side wall with tweeters inboard. The Rooze setup is clearly a radical idea and I have tried the Rooze with tweeters closer to the wall and away from the wall. I believe it works better with them closer to the wall especially with Dirac because the distance from the tweeter to the reflective wall is cut in half. I have not experimented with magnets flipped. As you can see in the diagram I have added 12" baffles (1" thick oak) to each side of the speaker which helps block the rear waves and gives a little bonus mid-bass boost. |
|
|
|
Post by leonski on Feb 5, 2021 19:39:43 GMT -5
I've not experimented with the rooze setup. Well aware, however. don't forget that they DON'T radiate to the 'sides' so your distance idea may not be the best model of what's going on.
People have added 'wings' to panels for years. The front/ back cancelation at the EDGE OF THE PANEL can be changed with such wings.
As it turns out? The 20 series is push / pull not single ended like all the others. And the newest panel? The 30.7 is multiple panels PER SIDE just like the originals. But I don't know
if it is a single ended OR push pull design.
The Magnepna ribbon tweeter is considered one of the best of its kind available.
|
|
|
|
Post by marcl on Feb 6, 2021 2:22:41 GMT -5
I've not experimented with the rooze setup. Well aware, however. don't forget that they DON'T radiate to the 'sides' so your distance idea may not be the best model of what's going on. People have added 'wings' to panels for years. The front/ back cancelation at the EDGE OF THE PANEL can be changed with such wings. As it turns out? The 20 series is push / pull not single ended like all the others. And the newest panel? The 30.7 is multiple panels PER SIDE just like the originals. But I don't know if it is a single ended OR push pull design. The Magnepna ribbon tweeter is considered one of the best of its kind available. To stay on topic (Dirac) ... I posted my rather unconventional setup to encourage a few things: first, don't be afraid to experiment with mic positions; second, listen AND measure because doing a lot of both can lead you to incrementally better sound; third, if you hear a problem look for the source and consider measurements a tool to find a solution (i.e. it may not be Dirac's fault). All of this may be especially important with dipole speakers and/or very reflective or asymmetrical rooms. So in my case I heard a problem with imaging, measured delays and misalignment after Dirac calibration, used temporary absorbers and baffles to isolate the problem, changed speaker position and measured and calibrated again .... and so I determined that getting the tweeters closer to the wall reduced Dirac's propensity to get confused and misalign the fronts. I also came up with the idea to use temporary baffles only for the MLP measurement. One key to the Rooze setup is aiming the dipole null at the MLP. My baffles extend the speakers to improve the effect. And the Rooze also sends the back wave on a very long journey across the room where I scatter it with lots of diffusion. There is no first reflection issue, because the first reflection IS the direct sound. Yes, the ribbon tweeters are excellent! And I'm very happy to have the same ribbons as the 30.7 and the (hopefully in production someday) 30.7 "for Condos". I've heard both of those and they are the best speakers I've ever heard. And BTW, notice that my setup uses two push-pull DWM's to augment the 3.7 bass. |
|
|
|
Post by hsamwel on Feb 7, 2021 15:51:06 GMT -5
Dirac and Levels There has been some confusion about how Processor Levels are affected or not by Dirac. So I did a test to point out what "zero" is before and after Dirac. Zero always remains zero. So Dirac does not change Levels, we do. The following shows what happens when Processor Levels are set to 0.0dB and -12dB, before and after Dirac filter is loaded. These are the settings I used for this test. View Attachment View AttachmentI set the Left speaker to -12dB and ran Dirac. This is a screenshot of the Volume Calibration screen in Dirac. Please note that while the Left speaker is set at -12dB in the XMC Levels, Dirac doesn't care and completely disregards the XMC setting. View AttachmentThis spreadsheet shows the actual measured output using the settings above. Notice that while Dirac apparently changes the Overall System Output by a tiny amount, it does not change the relationship between the channels as it relates to Levels Settings in the processor so -12dB remains a vastly lower output level and must be raised back to 0.0dB to match the other speaker. View AttachmentJust to be clear. Dirac does nothing - relatively speaking - to Levels. 0.0dB in Processor Levels Settings will ALWAYS be 0.0dB and is 0.0dB relative to other speakers. I think most of the confusion is about changing levels after calibration. Which I think in most cases are not recommended. Atleast not for changing just for preference. In those cases it is better to change the curve in Dirac. Adjusting for making levels even I’m not sure about. I would (I'm a racist) to hear what Dirac has to say about that before changing my levels. |
|
|
|
Post by hsamwel on Feb 7, 2021 15:57:52 GMT -5
Thanks, I guess it just not very user friendly then  I think Dirac don’t believe you ”need” to lower the whole curve, or even should. That’s why it lacks a ”move whole curve up/down” feature. IMO as long as you don’t have several major dips of more than -10db which Dirac aren’t able to ”fix” you shoudn’t bother with moving the curve. |
|
|
|
Post by hsamwel on Feb 7, 2021 16:15:34 GMT -5
I am thinking to get 2 more subwoofer. Currently I am using 2 subwoofer. Mostly for movies. Right now I have 2 options. I am using mini dsp. Just for example input 1 is sub 1 and 2 and input 2 is for subwoofer 3 and 4. And all 4 subwoofers already sync together using minidsp. My question is: Option 1. If I connected center sub to minidsp input 1 and set as LFE and right sub mono to input 2 minidsp, does the bass will still sync, because LFE is no bass management. Option 2 is: directly (without minidsp) connected using hi level input sub 3 to front left and sub 4 to front right, the idea is make my front speaker full range and set the front speaker as large, and center sub connected to input 1 minidsp and set as LFE. Is this ok? Cause I read many forums saying that if you set your front speakers as large it will messed up the low freq. Any idea which one is better? Thank you. In most cases having all speakes set to small is best. There can however be cases where it isn’t. But setting all small and a 80hz (or higher for those speakers not capable enough) crossover is the best general setting for most. Working from there you can try what fits your setup best. For example lowering the crossover for the fronts to 50-70hz. Although having several additional ”subs” means you can make havok with your (hopefully) finely tuned bass. Because the fronts bass aren’t calibrated with the subs as a mono channel. Atleast not until we get DLBM. Just like most don’t recommend using L/R sub, using L/R fronts for BM shouldn’t be recommended for the very same reason. Using large for 2ch is a little different case though.. But in theory a small+sub setup should sound better if the sub is good enough to keep up with the fronts. |
|
|
|
Post by hsamwel on Feb 7, 2021 16:48:20 GMT -5
That's interesting about the V shape. So having intermediate mic points in the "mouth" of the V caused some issues? Also, must be nice to have such a quiet room! Right now, with fridge and HVAC off, the noise floor is 45dB(C) according to REW's SPL app. About 43dbC at the MLP at the moment but sometimes the refrigerator makes a couple db more noise. I always see it in the Waterfall at 40Hz. With HVAC and refrigerator off I pick up another 3-4db but I don't find it makes any difference to the measurements. Moving the computers and NAS boxes away from the MLP I was able to pick up about 6db. The reason for the tight arrangement and the V is to keep the mic inside the envelope of the shape of the angle of the fronts, not outside. The actual direct sound of the fronts that I listen to is bounced off the side walls, but if I get the mic too far left and right of MLP it also hears some of the direct (not bounced) sound from the side of the speakers facing the walls. This causes Dirac to get extraneous sound 1-2ms out of sync that does not contribute well to the listening at the MLP. When I had the speakers in a conventional front-facing and toed in arrangement, I used a wider rectangular pattern across both cushions on the couch, with good results. That’s why it’s better to lower the mic gain and raise the main volume. But keep it so it won’t play too loud of course. If you have a high noise floor that is.. |
|
