|
|
Post by cullenhall on Jun 5, 2017 13:41:49 GMT -5
If this has been asked, let alone answered, I sure cant find it.
I very much enjoy Dirac and results that I am hearing, I am still tweaking the target curves to get something that I like in my spaces.
It occurs to me that I am BLIND as to what Dirac is doing in terms of the parameters that it applies to the various filters. so to me that begs the question, is there anyway to see what filters that DIRAC designs? I can see the before and the after, however it does not tell me exactly what filters it is applying. I understand that DIRAC is applying filters that go beyond simple parametric EQs, however, I am sure after some study one could understand how the parameters cause the filters to perform.
Anyone know how to do this?
|
|
|
|
Post by geebo on Jun 5, 2017 14:18:22 GMT -5
If this has been asked, let alone answered, I sure cant find it. I very much enjoy Dirac and results that I am hearing, I am still tweaking the target curves to get something that I like in my spaces. It occurs to me that I am BLIND as to what Dirac is doing in terms of the parameters that it applies to the various filters. so to me that begs the question, is there anyway to see what filters that DIRAC designs? I can see the before and the after, however it does not tell me exactly what filters it is applying. I understand that DIRAC is applying filters that go beyond simple parametric EQs, however, I am sure after some study one could understand how the parameters cause the filters to perform. Anyone know how to do this? That is information that only Dirac knows. There is no way for you to see the actual filter parameters applied. |
|
|
|
Post by cullenhall on Jun 5, 2017 14:56:03 GMT -5
I thought this might be the case. I can see how they would want to protect whatever "secret sauce" they have come up with - cant blame them one bit. I suppose (at least in my mind) this is a reason to apply filters by hand or with the help of REW - at least then I can look at what the software calculates and force it to do something that I might prefer.
|
|
|
|
Post by Axis on Jun 5, 2017 15:12:12 GMT -5
If this has been asked, let alone answered, I sure cant find it. I very much enjoy Dirac and results that I am hearing, I am still tweaking the target curves to get something that I like in my spaces. It occurs to me that I am BLIND as to what Dirac is doing in terms of the parameters that it applies to the various filters. so to me that begs the question, is there anyway to see what filters that DIRAC designs? I can see the before and the after, however it does not tell me exactly what filters it is applying. I understand that DIRAC is applying filters that go beyond simple parametric EQs, however, I am sure after some study one could understand how the parameters cause the filters to perform. Anyone know how to do this? That is information that only Dirac knows. There is no way for you to see the actual filter parameters applied. Lonnie showed me some graphs at the Atlanta event. I thought that you could. If you cant then maybe it is for technical reason. All that matters is the results. Myself I would just let Dirac sort it out. The folks that made it went deep into the woods so I do not have to. You can play with any adjustment that allows you to. Take a break and let technology do it's job and if it needs a tweak use the tools that come with it or get your own and adjust away. I would not consider not knowing the final filters for a very complicated software solution to be a fault in it. That may be a future feature that will come out from those who are in charge of it. Emotiva is not the only one that uses Dirac but they have been involved since the beginning of Dirac. It is pretty cool that Emotiva has Dirac room correction and had input since the beginning. The other gear that has it cost as much as a good car. Emotiva made it work in the XMC-1 brilliantly and made it a hard excuse to buy the more costly gear when the XMC-1 nails it. The last show blew the socks off folks with the T2's and S15's and barrage of supporting speakers. I think the secrete sauce is the supporting analog system along with the excellent room correction. Can it be improved on ? Yes |
|
klinemj
Emo VIPs 
Official Emofest Scribe
Posts: 15,088
|
Post by klinemj on Jun 5, 2017 15:26:15 GMT -5
I thought this might be the case. I can see how they would want to protect whatever "secret sauce" they have come up with - cant blame them one bit. I suppose (at least in my mind) this is a reason to apply filters by hand or with the help of REW - at least then I can look at what the software calculates and force it to do something that I might prefer. I believe you can find some white papers on what Dirac does...if I recall correctly, they used to even have links to them on their site. Check around. While it's fairly high end stuff, if you have ever studied process control theory and designed a controller, it will seem fairly clear to you. If not, the math gets pretty deep. I'm a Chemical Engineer so I studied that stuff (as it relates to controlling processes of making chemicals, but the theory and math are essentially the same), had to design my own controllers for classes, and that taught me...I'd rather do something different than process controls for my career. That said...even if you cannot understand what Dirac is doing, if you like the results - just roll with it. With the filters of Dirac, if you were to try to manually override the parameters in the filters without understanding control theory, odds are you'd just make the result worse (trust me, I know...from my Process Controls class before and after I understood the theory!). Mark |
|
|
|
Post by geebo on Jun 5, 2017 15:46:26 GMT -5
I thought this might be the case. I can see how they would want to protect whatever "secret sauce" they have come up with - cant blame them one bit. I suppose (at least in my mind) this is a reason to apply filters by hand or with the help of REW - at least then I can look at what the software calculates and force it to do something that I might prefer. I believe you can find some white papers on what Dirac does...if I recall correctly, they used to even have links to them on their site. Check around. While it's fairly high end stuff, if you have ever studied process control theory and designed a controller, it will seem fairly clear to you. If not, the math gets pretty deep. I'm a Chemical Engineer so I studied that stuff (as it relates to controlling processes of making chemicals, but the theory and math are essentially the same), had to design my own controllers for classes, and that taught me...I'd rather do something different than process controls for my career. That said...even if you cannot understand what Dirac is doing, if you like the results - just roll with it. With the filters of Dirac, if you were to try to manually override the parameters in the filters without understanding control theory, odds are you'd just make the result worse (trust me, I know...from my Process Controls class before and after I understood the theory!). Mark Yeah, the best way to affect a change in the Dirac filters is to simply adjust the target curves. |
|
|
|
Post by cullenhall on Jun 5, 2017 15:47:07 GMT -5
All good points guys, I am an electrical engineer and have a minor in computer science, so I think at the end of the day "black boxes" to me are just begging to be opened up and understood! I have to say I really do like the results that I am getting and I am sure given enough time I will find a house curve that works for me and suits my taste (and slight high end hearing loss!) and I'll be as pleased as can be.
Many years ago I owned a touring sound company and to date still own a recording studio (one of the few end to end workflow DSD studios) and have always done my speaker calibrations by hand till the Meyer SIM2 came along. Lots of great stuff in the SIM2 but in the end I still saw all the "massaging" that it was doing to my signal so that gave me "comfort". The DIRAC is as good or better than the SIM2 but I cant see what its doing - so I guess that makes me skeptical.
Thanks for listening to my musings on this subject - all fascinating stuff!
|
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,088
|
Post by klinemj on Jun 5, 2017 17:05:28 GMT -5
|
|
|
|
Post by wilburthegoose on Jun 5, 2017 17:40:37 GMT -5
I found the default house curves to beat my own curves 100% of the time.
|
|
|
|
Post by Axis on Jun 5, 2017 17:52:30 GMT -5
The curves that folks were adjusting away from the house curves was based on a peculiar (strange or odd; unusual) curve that would benefit there system. The house curve is a house curve that the folks that brought you the XMC-1 and Dirac decided was the best curve for most. If you are going to mess the black box and break into the adjustments you have to earn it.
|
|
|
|
Post by cullenhall on Jun 5, 2017 19:37:09 GMT -5
I found the default house curves to beat my own curves 100% of the time. It's interesting that you say that, at my lakehouse (xmc-1 and SVS Ultra speakers) I liked the default curve immensely and has left me with no desire to change it. In my main home (xmc-1 and Emotiva Airmotiv speakers) where I have a far more treated and dedicated space I found the default curve to be thin on the bottom end and perhaps a bit bright. Probably something that I need to dig into in ways other that just tweaking house curves. |
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,088
|
Post by klinemj on Jun 5, 2017 20:15:34 GMT -5
I found the default house curves to beat my own curves 100% of the time. It's interesting that you say that, at my lakehouse (xmc-1 and SVS Ultra speakers) I liked the default curve immensely and has left me with no desire to change it. In my main home (xmc-1 and Emotiva Airmotiv speakers) where I have a far more treated and dedicated space I found the default curve to be thin on the bottom end and perhaps a bit bright. Probably something that I need to dig into in ways other that just tweaking house curves. I suspect the differences between your sites has more to do with differences in what the system read (driven by mic placements and level setting) than anything else. I would re-run the main house system a few times and see what happens. Mark |
|
|
|
Post by cullenhall on Jun 5, 2017 20:21:40 GMT -5
It's interesting that you say that, at my lakehouse (xmc-1 and SVS Ultra speakers) I liked the default curve immensely and has left me with no desire to change it. In my main home (xmc-1 and Emotiva Airmotiv speakers) where I have a far more treated and dedicated space I found the default curve to be thin on the bottom end and perhaps a bit bright. Probably something that I need to dig into in ways other that just tweaking house curves. I suspect the differences between your sites has more to do with differences in what the system read (driven by mic placements and level setting) than anything else. I would re-run the main house system a few times and see what happens. Mark I'll keep messing with it that's for sure - with 4 kids in the house it's hard to find gobs of time to sit and tweak things a bunch. |
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,088
|
Post by klinemj on Jun 5, 2017 20:25:03 GMT -5
I suspect the differences between your sites has more to do with differences in what the system read (driven by mic placements and level setting) than anything else. I would re-run the main house system a few times and see what happens. Mark I'll keep messing with it that's for sure - with 4 kids in the house it's hard to find gobs of time to sit and tweak things a bunch. I hear you on that...despite me telling the wife and kids "I am doing something special...you will hear odd noises...ignore them and be quiet"...they do not. They come running every time and say "what are you doing?" Every. Single. Time. Which leads to a discussion and sending each away. It's best for me to run it when they are not around. Mark |
|
|
|
Post by qdtjni on Jun 6, 2017 2:08:43 GMT -5
This White Paper gives some more insight, diracdocs.com/on_room_correction.pdfAlso, some the academic papers authored or co-authored by Dr. Mathias Johansson from his time at Uppsala University may give even more insight. |
|
