|
|
Post by nickwin on Mar 30, 2015 7:52:42 GMT -5
I ran Dirac for the first time today (full version). It sounds really good in a lot of ways, but the bass, lower mids seemed off to me. Ran a couple quick measurements with REW and this is what I got. I set up my subwoofer target curve to rise 5db between 80 and 20 hz. My cross over is 100hz. To me it looks like Dirac over compensated in the 100-300hz range and set the subwoofer level to low. Whats going on, should I remeasure? The first graph is dirac exactly how it came out and the second is with 5.5db added to the sub. Have you tried tweaking the distance to see if you can find a smoother response at the crossover point? If you smooth out the crossover before running Dirac, it might not have to compensate as much. That said, it doesn't look horrible to me after you boosted the sub level. How does it sound with the +5.5? Also, have you used and SPL meter to verify the levels? I found that after running Dirac LE (still waiting on the key for the full version), my speakers were not level matched, with the biggest problem being my sub. Edit: I generally use the SPL meter within REW rather than my RS meter since I can use a calibrated mic with REW. Edit #2: I just remembered that distance is removed when you run Dirac.  To me it sounded ok with the bass at +5.5 but I felt the tonality was off. There was a ton of mid bass that just didn't sound natural to me and even at +5.5 it drown out the lower bass. I switched to socketmans vs file and it sounds MUCH better to my ears. The graphs its producing also look much more linear. I highly recommend using an accurate 90° cal over Emotivas. My newest graphs with sockemans cal file look super linear BUT I'm having trouble with the bass that's routed through the crossovers. If you could play with the distance setting post Dirac it would be really nice. |
|
|
|
Post by bluescale on Mar 30, 2015 16:22:16 GMT -5
To me it sounded ok with the bass at +5.5 but I felt the tonality was off. There was a ton of mid bass that just didn't sound natural to me and even at +5.5 it drown out the lower bass. I switched to socketmans vs file and it sounds MUCH better to my ears. The graphs its producing also look much more linear. I highly recommend using an accurate 90° cal over Emotivas. Funny you should say that. When I first got my Dirac LE license, I tried the measurement for a few days, and decided to ditch it and go back to my manual REW EQ. I was disappointed because I'd used the trial version of Dirac last summer on my HTPC, and LOVED what it did for my system. I was very disappointed with the broken mic calibration that comes with Dirac LE. After someone told me about the cal file available here, I tried that, and it got me very close. That's what I'm using currently. Hopefully tonight I'll get the chance to use my CSL calibrated UMIK-1 and see how it does. Have you tried changing the crossover to see what happens if you move it up or down on either side of 100hz? It's a bit of work, but it might result in a smoother transition. I believe the DataSat processor allows you to adjust distance after running Dirac. I wish the XMC-1 would allow that as well. It would be a great feature, especially for folks with multiple subs. |
|
|
|
Post by nickwin on Mar 30, 2015 16:36:21 GMT -5
To me it sounded ok with the bass at +5.5 but I felt the tonality was off. There was a ton of mid bass that just didn't sound natural to me and even at +5.5 it drown out the lower bass. I switched to socketmans vs file and it sounds MUCH better to my ears. The graphs its producing also look much more linear. I highly recommend using an accurate 90° cal over Emotivas. Funny you should say that. When I first got my Dirac LE license, I tried the measurement for a few days, and decided to ditch it and go back to my manual REW EQ. I was disappointed because I'd used the trial version of Dirac last summer on my HTPC, and LOVED what it did for my system. I was very disappointed with the broken mic calibration that comes with Dirac LE. After someone told me about the cal file available here, I tried that, and it got me very close. That's what I'm using currently. Hopefully tonight I'll get the chance to use my CSL calibrated UMIK-1 and see how it does. Have you tried changing the crossover to see what happens if you move it up or down on either side of 100hz? It's a bit of work, but it might result in a smoother transition. I believe the DataSat processor allows you to adjust distance after running Dirac. I wish the XMC-1 would allow that as well. It would be a great feature, especially for folks with multiple subs. Iv tried every crossover setting possible. the issue is my mains and subs are simply summing very poorly and there effecting each other all the way down to the bottom of by subs response which is 20hz. If I move the crossover point even 10hz either way I can optimize one area, say around the xover, but it will wildly throw off another area, say the lower bass. No matter what I do I cannot get the bass routed from my mains to be as linear as the sub on its own. And unfortunately its a big difference. See my thread about the crossover issue I'm having. |
|
|
|
Post by fbczar on Apr 1, 2015 7:11:48 GMT -5
To me it sounded ok with the bass at +5.5 but I felt the tonality was off. There was a ton of mid bass that just didn't sound natural to me and even at +5.5 it drown out the lower bass. I switched to socketmans vs file and it sounds MUCH better to my ears. The graphs its producing also look much more linear. I highly recommend using an accurate 90° cal over Emotivas. Funny you should say that. When I first got my Dirac LE license, I tried the measurement for a few days, and decided to ditch it and go back to my manual REW EQ. I was disappointed because I'd used the trial version of Dirac last summer on my HTPC, and LOVED what it did for my system. I was very disappointed with the broken mic calibration that comes with Dirac LE. After someone told me about the cal file available here, I tried that, and it got me very close. That's what I'm using currently. Hopefully tonight I'll get the chance to use my CSL calibrated UMIK-1 and see how it does. Have you tried changing the crossover to see what happens if you move it up or down on either side of 100hz? It's a bit of work, but it might result in a smoother transition. I believe the DataSat processor allows you to adjust distance after running Dirac. I wish the XMC-1 would allow that as well. It would be a great feature, especially for folks with multiple subs. bluscale, How have your results been with the UMIC-1 and the Cross Spectrum calibration files? Is the UMIK-1 with the Cross Spectrum calibration superior to the Emotiva microphone with Dirac Full? |
|
|
|
Post by bluescale on Apr 1, 2015 22:08:54 GMT -5
bluscale, How have your results been with the UMIC-1 and the Cross Spectrum calibration files? Is the UMIK-1 with the Cross Spectrum calibration superior to the Emotiva microphone with Dirac Full? Yes, the CSL calibrate umik-1 is definitely superior to the cal file provided by Emotiva for in the free version of Dirac Live. I haven't tried the new cal file, but indications are that it's good quality, but not perfect (what you'd expect from batch calibration). Socketman's cal file got me close previously, but using someone else's cal file is the equivalent of batch calibration. Things definitely sounds better now. Of course, that might be more a feature of an adjustable curve than my calibrated mic. Here's my thought process - we've all purchased a luxury item in the XMC-1. I'm not saying we're all right (I'm smack in the middle of middle class). However, those of us who have an XMC-1 can probably afford the little bit extra to get a calibrated microphone. For me it was worth taking the doubt and second guessing out of it to have a calibrated mic on hand. |
|
|
|
Post by fbczar on Apr 1, 2015 22:16:59 GMT -5
bluscale, How have your results been with the UMIC-1 and the Cross Spectrum calibration files? Is the UMIK-1 with the Cross Spectrum calibration superior to the Emotiva microphone with Dirac Full? Yes, the CSL calibrate umik-1 is definitely superior to the cal file provided by Emotiva for in the free version of Dirac Live. I haven't tried the new cal file, but indications are that it's good quality, but not perfect (what you'd expect from batch calibration). Socketman's cal file got me close previously, but using someone else's cal file is the equivalent of batch calibration. Things definitely sounds better now. Of course, that might be more a feature of an adjustable curve than my calibrated mic. Here's my thought process - we've all purchased a luxury item in the XMC-1. I'm not saying we're all right (I'm smack in the middle of middle class). However, those of us who have an XMC-1 can probably afford the little bit extra to get a calibrated microphone. For me it was worth taking the doubt and second guessing out of it to have a calibrated mic on hand. I am trying to use a 3 meter Amazon Basics USB extension cable with my UMIK-1, but Dirac Full and my IMac will not recognize the microphone when I use the extension cable. It will see the UMIK-1 when I connect it with the short USB cable supplied with the UMIK-1. Are you using an extension cable? Any suggestions? |
|
|
|
Post by bluescale on Apr 1, 2015 22:32:11 GMT -5
Yes, the CSL calibrate umik-1 is definitely superior to the cal file provided by Emotiva for in the free version of Dirac Live. I haven't tried the new cal file, but indications are that it's good quality, but not perfect (what you'd expect from batch calibration). Socketman's cal file got me close previously, but using someone else's cal file is the equivalent of batch calibration. Things definitely sounds better now. Of course, that might be more a feature of an adjustable curve than my calibrated mic. Here's my thought process - we've all purchased a luxury item in the XMC-1. I'm not saying we're all right (I'm smack in the middle of middle class). However, those of us who have an XMC-1 can probably afford the little bit extra to get a calibrated microphone. For me it was worth taking the doubt and second guessing out of it to have a calibrated mic on hand. I am trying to use a 3 meter Amazon Basics USB extension cable with my UMIK-1, but Dirac Full and my IMac will not recognize the microphone when I use the extension cable. It will see the UMIK-1 when I connect it with the short USB cable supplied with the UMIK-1. Are you using an extension cable? Any suggestions? I have not used an extension cable. I'm fortunate that my HTPC is close enough to my couch that have *just* enough slack to get to every measurement spot. An iMac is pretty easy to move around. While a bit annoying, can you just move it out a few feet, or will that get in the way of you measurements? |
|
|
|
Post by fbczar on Apr 2, 2015 1:43:11 GMT -5
I switched USB ports and now it works!
|
|
|
|
Post by lbrown105 on Apr 30, 2015 21:55:16 GMT -5
Dirac Live PC version always had a clip indicator light in the audio processor controller. How does this work with XMC-1 full version. I know it shows clipping in the measurement process but I am not talking about that, I am talking about after calibration while playing content?
|
|
|
|
Post by jclangston on May 4, 2015 15:10:08 GMT -5
Is it possible to have different Dirac settings for each input? The reason I ask is because it sounds different between HDMI 1 (Directv) and HDMI 2 (PS4). I have the PS4 set to bit stream output. It seems like the highs roll off more when watching a movie compared to watching Directv. There is a noticeable difference in EQ. I didn't know if I had changed a setting that I didn't know about...
|
|
|
|
Post by geebo on May 4, 2015 16:09:30 GMT -5
Is it possible to have different Dirac settings for each input? The reason I ask is because it sounds different between HDMI 1 (Directv) and HDMI 2 (PS4). I have the PS4 set to bit stream output. It seems like the highs roll off more when watching a movie compared to watching Directv. There is a noticeable difference in EQ. I didn't know if I had changed a setting that I didn't know about... Dirac is a really just another Preset and only creates one set of filters. Do you have the Dirac preset chosen for the different inputs? Each input can have any one of the three presets assigned to it as a default. |
|
ron82
Minor Hero
Posts: 12 
|
Post by ron82 on Jun 17, 2015 4:40:28 GMT -5
Not sure if this is the right place to post this. I have started to play around with DIRAC, and although I can hear the differences I'm not sure I like it. I have done some testing before and after with REW, but doing this I notice some very strange measurements. Using DIRAC and REW, and going for 256k length the curves is very close. But when doing a more detailed 1M length test in REW I have some large "valleys" in my frequency response. The biggest one is at 700 Hz. It starts going down at 500, reaching the bottom at ~650-700 Hz and back up at around 900-1000 Hz. Thee drop is over 20 dB, and it's there no matter if engage DIRAC or not. If i use 256k lengt in REW, this "valley is not coming up, it's actually a tiny spike there instead... I have also tried moving the speakers, not much help. I started playing some test tones at 500-800 Hz range, and just walking around the room, and some frequencies have large variations. Did the same with my mic. Had readings that differed up to 20 dB. Also noticed that when I found a frequency that would drop my dB reading down to low 60 in my sweet spot, just moving myself a few feet, or angling my laptop differently could spike the reading up to above 70 dB. I was positioned out to the side,standing in the room and holding my laptop, out of the way between the speakers and my sofa/sweetspot/mic. I have read a lot about basstraps, how to maybe reduce the spikes etc, or using different absorbers to reduce echo and stuff. But I have a very hard time find any info on how to deal with those large valleys I have. Is it due to standing waves and cancelling out? Every place I read they only look at this as a problem up to around 300 Hz. Under are attached REW curves.    |
|
|
|
Post by millst on Jun 17, 2015 10:44:43 GMT -5
You might be better off going to the HTS forums, where there is an entire section dedicated to REW.
I don't think there should be much of a difference between the different lengths (256K vs 1M). The advantage to the longer sweep is more signal-to-noise ratio; about 3dB per jump. Some room modes might appear worse since they have more time to build up, but your two are very different after 300Hz. Are you sure you aren't changing something else?
Any time you change the mic position at all, even inches, you may get very different measurements, which is normal. That's why Dirac takes multiple measurements and recommends you take the max number. It averages corrections across the entire area.
Room treatment is more effective than room correction, because the treatment will be effective for all positions. Any room correction system will have to make a compromise on optimal correction vs target area. You could place all the measurement points in a cube where your head would be at the MLP, but then all the other seats get worse response. Room treatment isn't something to be taken lightly. You can't just slap some stuff on the wall, however.
-tm
|
|
Chris
Emo VIPs 
Posts: 424
|
Post by Chris on Jul 3, 2015 17:23:27 GMT -5
Hi,
After having the XMC-1 for a few months, I finally decided to give Direc LE a try. I have never been a fan of previous automated calibrated systems and never bothered to use the EMO-Q in my previous UMC-1. I just set things up manually and got reasonable results.
I have now run Dirac LE four times. Each time I believe I made improvements in the capture process as I learned what Dirac LE was looking for in levels. On my fourth run, I got very consistent results in the levels captured based on the wave form uniformity I saw as it was recorded by Direc LE.
The resulting adjustments that Direc LE applied seemed reasonable except for the subwoofer adjustment which I believe is at least 10 db lower than what is correct for my listening room. I notice that Kal Rubinson in this recent Stereophile review also seemed to take issue with the subwoofer adjustment: "My run of Dirac LE was not a great success: It sharply reduced the subwoofers' outputs". This seems to my case as well. The easy fix is simply to go in to the levels and change the subwoofer level in the Dirac preset by increasing 10 db. Or, I could change the level control on my sub (ML Grotto i) which is currently at about 1'oclock but that would mess up my capture level if I run Dirac LE again. Kal later installs the full Dirac and seems to get better results (although he apparently changed the main speakers mid-stream which adds another variable?). I also think I have seen other comments on this board about the subwoofer level after running Dirac LE?
My question is what is the potential reason the sub level is reduced too much by Dirac? Also, why did the full Dirac give different results? My understanding of the differences between LE and Full is that Full allows various parameters to be adjusted but I would think the basic measurement engine would be the same and give the same results?
I seem to remember way back that there was discussion about sub levels in the original UMC-1 and EMO-Q which may have received a firmware update to address? My memory is hazy on this.
Anyway I would love any feedback on Dirac LE and Sub levels. I am virtually certain that after running Dirac LE the Sub is too low in actual use and the graphs in Dirac LE also reflect a reduction in Sub levels that I am hearing.
|
|
|
|
Post by geebo on Jul 3, 2015 17:53:15 GMT -5
Hi, After having the XMC-1 for a few months, I finally decided to give Direc LE a try. I have never been a fan of previous automated calibrated systems and never bothered to use the EMO-Q in my previous UMC-1. I just set things up manually and got reasonable results. I have now run Dirac LE four times. Each time I believe I made improvements in the capture process as I learned what Dirac LE was looking for in levels. On my forth run, I got very consistent results in the levels captured based on the wave form uniformity I saw as it was recorded by Direc LE. The resulting adjustments that Direc LE applied seemed reasonable except for the subwoofer adjustment which I believe is at least 10 db lower than what is correct for my listening room. I notice that Kal Rubinson in this recent Stereophile review also seemed to take issue with the subwoofer adjustment: "My run of Dirac LE was not a great success: It sharply reduced the subwoofers' outputs". This seems to my case as well. The easy fix is simply to go in to the levels and change the subwoofer level in the Dirac preset by increasing 10 db. Or, I could change the level control on my sub (ML Grotto i) which is currently at about 1'oclock but that would mess up my capture level if I run Dirac LE again. Kal later installs the full Dirac and seems to get better results (although he apparently changed the main speakers mid-stream which adds another variable?). I also think I have seen other comments on this board about the subwoofer level after running Dirac LE? My question is what is the potential reason the sub level is reduced too much by Dirac? Also, why did the full Dirac give different results? My understanding of the differences between LE and Full is that Full allows various parameters to be adjusted but I would think the basic measurement engine would be the same and give the same results? I seem to remember way back that there was discussion about sub levels in the original UMC-1 and EMO-Q which may have received a firmware update to address? My memory is hazy on this. Anyway I would love any feedback on Dirac LE and Sub levels. I am virtually certain that after running Dirac LE the Sub is too low in actual use and the graphs in Dirac LE also reflect a reduction in Sub levels that I am hearing. The main reason for the big differences between Dirac LE and Dirac Full was that he also changed mic and calibration file. With LE, I also get reduced bass but with my UMIK and Dirac Full I get a much better and pleasing balance especially at the bottom end. |
|
Chris
Emo VIPs
Posts: 424
|
Post by Chris on Jul 4, 2015 11:16:59 GMT -5
With LE, I also get reduced bass but with my UMIK and Dirac Full I get a much better and pleasing balance especially at the bottom end. Has anyone from Emotiva commented on these inconsistent results? Is it something that might be addressed in a future firmware upgrade? Automated calibration like Dirac has the potential to be great but if there are inconsistencies and potential errors like some of us are experiencing with sub levels it seeds doubt and confusion. Obviously, I can just go back to my manual setup in my Preset 1, but I was hoping that by harnessing the power of an external computer that automated calibration had reached a new level accuracy and consistency? This is really confusing to me as I would've thought Dirac LE would have at the very least gotten the bass calibration correct? I believe the bass area is the most problematic for most home users? Adjustment of higher frequencies is icing on the cake but I consider getting the bass correct a critical feature? -CB |
|
|
|
Post by geebo on Jul 4, 2015 11:39:01 GMT -5
With LE, I also get reduced bass but with my UMIK and Dirac Full I get a much better and pleasing balance especially at the bottom end. Has anyone from Emotiva commented on these inconsistent results? Is it something that might be addressed in a future firmware upgrade? Automated calibration like Dirac has the potential to be great but if there are inconsistencies and potential errors like some of us are experiencing with sub levels it seeds doubt and confusion. Obviously, I can just go back to my manual setup in my Preset 1, but I was hoping that by harnessing the power of an external computer that automated calibration had reached a new level accuracy and consistency? This is really confusing to me as I would've thought Dirac LE would have at the very least gotten the bass calibration correct? I believe the bass area is the most problematic for most home users? Adjustment of higher frequencies is icing on the cake but I consider getting the bass correct a critical feature? -CB The full power of Dirac is realized with the upgrade for 99 bucks and was one of the best 99 bucks I've spent. LE was consistent for me but it was consistently weak in the bass and I had to compensate with the bass controls and or playing with the calibration files. Other users seem to like LE just fine. I got the full version along with an individually calibrated mic and in the process got by far the best room calibration I've ever experienced and in fact the only one I've preferred over manual correction. The full version really unleashes the power of Dirac. |
|
LCSeminole
Global Moderator 
Res firma mitescere nescit.
Posts: 20,850
Member is Online
|
Post by LCSeminole on Jul 4, 2015 12:17:30 GMT -5
The full power of Dirac is realized with the upgrade for 99 bucks and was one of the best 99 bucks I've spent. LE was consistent for me but it was consistently weak in the bass and I had to compensate with the bass controls and or playing with the calibration files. Other users seem to like LE just fine. I got the full version along with an individually calibrated mic and in the process got by far the best room calibration I've ever experienced and in fact the only one I've preferred over manual correction. The full version really unleashes the power of Dirac. Across the board, these are my results as well. Besides maybe room treatments, the $99 upgrade is so far, by far the deal of this decade for me. |
|
|
|
Post by ansat on Jul 11, 2015 10:45:21 GMT -5
Hi, After having the XMC-1 for a few months, I finally decided to give Direc LE a try. I have never been a fan of previous automated calibrated systems and never bothered to use the EMO-Q in my previous UMC-1. I just set things up manually and got reasonable results. I have now run Dirac LE four times. Each time I believe I made improvements in the capture process as I learned what Dirac LE was looking for in levels. On my fourth run, I got very consistent results in the levels captured based on the wave form uniformity I saw as it was recorded by Direc LE. The resulting adjustments that Direc LE applied seemed reasonable except for the subwoofer adjustment which I believe is at least 10 db lower than what is correct for my listening room. I notice that Kal Rubinson in this recent Stereophile review also seemed to take issue with the subwoofer adjustment: "My run of Dirac LE was not a great success: It sharply reduced the subwoofers' outputs". This seems to my case as well. The easy fix is simply to go in to the levels and change the subwoofer level in the Dirac preset by increasing 10 db. Or, I could change the level control on my sub (ML Grotto i) which is currently at about 1'oclock but that would mess up my capture level if I run Dirac LE again. Kal later installs the full Dirac and seems to get better results (although he apparently changed the main speakers mid-stream which adds another variable?). I also think I have seen other comments on this board about the subwoofer level after running Dirac LE? My question is what is the potential reason the sub level is reduced too much by Dirac? Also, why did the full Dirac give different results? My understanding of the differences between LE and Full is that Full allows various parameters to be adjusted but I would think the basic measurement engine would be the same and give the same results? I seem to remember way back that there was discussion about sub levels in the original UMC-1 and EMO-Q which may have received a firmware update to address? My memory is hazy on this. Anyway I would love any feedback on Dirac LE and Sub levels. I am virtually certain that after running Dirac LE the Sub is too low in actual use and the graphs in Dirac LE also reflect a reduction in Sub levels that I am hearing. Hey Chris, I have a whole slew of stuff in the forums for weak bass / harsh highs. The issue I found was that on LE - The calibration file is not true. The official response from Lonnie was that the calibration file and Dirac LE curve were adjusted to work together to come up with what they found to be a pleasing sound. Simple put, the measurements your seeing on Dirac with the emotiva mic are not accurate representation and your end result will not be the +3 to -3 as indicated by the dirac target curve. To correct this, I generated a calibration file from my cross spectrum microphones and this got things much closer. On top of that, Socketman sent his microphone to cross spectrum and shared his calibration. Both can be found in my signature along with instructions on how to apply them to Dirac LE. If you want to customize the curve further, I recommend the full version. (I still use my calibration files with the full version when using the EMM-1). Tony |
|
|
|
Post by ansat on Jul 11, 2015 10:55:12 GMT -5
Not sure if this is the right place to post this. I have started to play around with DIRAC, and although I can hear the differences I'm not sure I like it. I have done some testing before and after with REW, but doing this I notice some very strange measurements. Using DIRAC and REW, and going for 256k length the curves is very close. But when doing a more detailed 1M length test in REW I have some large "valleys" in my frequency response. The biggest one is at 700 Hz. It starts going down at 500, reaching the bottom at ~650-700 Hz and back up at around 900-1000 Hz. Thee drop is over 20 dB, and it's there no matter if engage DIRAC or not. If i use 256k lengt in REW, this "valley is not coming up, it's actually a tiny spike there instead... I have also tried moving the speakers, not much help. I started playing some test tones at 500-800 Hz range, and just walking around the room, and some frequencies have large variations. Did the same with my mic. Had readings that differed up to 20 dB. Also noticed that when I found a frequency that would drop my dB reading down to low 60 in my sweet spot, just moving myself a few feet, or angling my laptop differently could spike the reading up to above 70 dB. I was positioned out to the side,standing in the room and holding my laptop, out of the way between the speakers and my sofa/sweetspot/mic. I have read a lot about basstraps, how to maybe reduce the spikes etc, or using different absorbers to reduce echo and stuff. But I have a very hard time find any info on how to deal with those large valleys I have. Is it due to standing waves and cancelling out? Every place I read they only look at this as a problem up to around 300 Hz. Under are attached REW curves. If you are using the EMM-1, this microphone does not like anything besides 256. I think this might be a limitation of the USB interface. I would not trust the results with anything above that. Tony |
|
