|
|
Post by hsamwel on Dec 5, 2020 8:18:21 GMT -5
Why do you extend your sub out of its capabilities? To me it seems 20hz at 0db is were the drop should be. IMO you don’t need to steer the drop in the low end. Moving the curtain until you get it to fall naturally with the measured. Although I agree with the general point of your post. I do this myself for several speakers, among them my subwoofers. My sub is spec'ed to 17Hz and it measures that low. I sure don't listen to pipe organs that often, but I have measured the 16Hz 32' rank in the Saint Saens Organ Symphony. So just on principle I let Dirac flatten the response down that low. My point in widening the curtains is to use Dirac to limit the response of the subs and also the other speakers. When you pull the curtains in tight, the speakers are left to respond outside that range. By widening the curtains and tightening the target curves I explicitly limit the response to what the speakers can do without straining their limits. What I mean is what I marked out on your screenshot. To me it measures 0db to 20hz but yes 17hz with -4 to -6db.. I would have put the last flat point at 20hz and let it fall with the measured line but with a steering point like you’ve added. I get your point of using the curtains, I do use rhem this way myself on many of my speakers. Attachments:
|
|
|
|
Post by marcl on Dec 5, 2020 8:33:11 GMT -5
My sub is spec'ed to 17Hz and it measures that low. I sure don't listen to pipe organs that often, but I have measured the 16Hz 32' rank in the Saint Saens Organ Symphony. So just on principle I let Dirac flatten the response down that low. My point in widening the curtains is to use Dirac to limit the response of the subs and also the other speakers. When you pull the curtains in tight, the speakers are left to respond outside that range. By widening the curtains and tightening the target curves I explicitly limit the response to what the speakers can do without straining their limits. What I mean is what I marked out on your screenshot. To me it measures 0db to 20hz but yes 17hz with -4 to -6db.. I would have put the last flat point at 20hz and let it fall with the measured line but with a steering point like you’ve added. I get your point of using the curtains, I do use rhem this way myself on many of my speakers. Aha okay I thought about that later. It has to do with that huge resonance at 40Hz. It responds to 16Hz before the steep rolloff, but I had to lower the sub level so Dirac wouldn't clip. Think of it as though 16Hz is 0db, and Dirac is pulling that peak down. It works okay with the target curve as I did it, but I also have done basically the same thing by lowering the target curve -6db, let Dirac pull the whole thing down, then pull the level back up +6db in the XMC or miniDSP. Here's a shot of all the measurements. My L/R also go to the subs below 35Hz with a Y connector to the miniDSP. Notice their response is above 0db at 16Hz, but Dirac measured the sub channel several db lower.  |
|
|
|
Post by hsamwel on Dec 5, 2020 11:45:01 GMT -5
What I mean is what I marked out on your screenshot. To me it measures 0db to 20hz but yes 17hz with -4 to -6db.. I would have put the last flat point at 20hz and let it fall with the measured line but with a steering point like you’ve added. I get your point of using the curtains, I do use rhem this way myself on many of my speakers. Aha okay I thought about that later. It has to do with that huge resonance at 40Hz. It responds to 16Hz before the steep rolloff, but I had to lower the sub level so Dirac wouldn't clip. Think of it as though 16Hz is 0db, and Dirac is pulling that peak down. It works okay with the target curve as I did it, but I also have done basically the same thing by lowering the target curve -6db, let Dirac pull the whole thing down, then pull the level back up +6db in the XMC or miniDSP. Here's a shot of all the measurements. My L/R also go to the subs below 35Hz with a Y connector to the miniDSP. Notice their response is above 0db at 16Hz, but Dirac measured the sub channel several db lower. View AttachmentOK. I still wouldn't extend it myself though.. But I see what you mean with it responding to 16hz. The little ladder..
Dirac themselves actually says it's ok to extend as long as you don't reach -8db to -10db.
Does Dirac measure them lower or do they repond lower in their place in the room? Maybe move them? Dirac should have each speaker at the same volume accrding to the curve.. Or?
Here's my measurements of the subs.. The left and right I have changed to totally flat.. The other on is the one I use with a Harman 4db curve..
|
|
|
|
Post by marcl on Dec 5, 2020 12:01:49 GMT -5
Aha okay I thought about that later. It has to do with that huge resonance at 40Hz. It responds to 16Hz before the steep rolloff, but I had to lower the sub level so Dirac wouldn't clip. Think of it as though 16Hz is 0db, and Dirac is pulling that peak down. It works okay with the target curve as I did it, but I also have done basically the same thing by lowering the target curve -6db, let Dirac pull the whole thing down, then pull the level back up +6db in the XMC or miniDSP. Here's a shot of all the measurements. My L/R also go to the subs below 35Hz with a Y connector to the miniDSP. Notice their response is above 0db at 16Hz, but Dirac measured the sub channel several db lower. View AttachmentOK. I still wouldn't extend it myself though.. But I see what you mean with it responding to 16hz. The little ladder..
Dirac themselves actually says it's ok to extend as long as you don't reach -8db to -10db.
Does Dirac measure them lower or do they repond lower in their place in the room? Maybe move them? Dirac should have each speaker at the same volume accrding to the curve.. Or?
Here's my measurements of the subs.. The left and right I have changed to totally flat.. The other on is the one I use with a Harman 4db curve..
I understand  When I do volume calibration in Dirac I always set all the speakers the same. My subs (two subs driven by a miniDSP) are always louder than the other speakers so I set them first. But for some reason Dirac still measured them several db lower than the main speakers. I've tried several ways of setting the sub level and the result is always the same. I even set the sub level first and set the other speakers a few db lower ... same result. I sent screen shots to Dirac and exchanged emails with Flavio and never got an answer to that issue. He just said as long as Dirac doesn't throw a clipping error or low signal error the measurements are okay, So I measure in REW after and sometimes I raise or lower the sub level for a smooth transition. And I always use flat targets. |
|
|
|
Post by hsamwel on Dec 5, 2020 12:27:48 GMT -5
OK. I still wouldn't extend it myself though.. But I see what you mean with it responding to 16hz. The little ladder..
Dirac themselves actually says it's ok to extend as long as you don't reach -8db to -10db.
Does Dirac measure them lower or do they repond lower in their place in the room? Maybe move them? Dirac should have each speaker at the same volume accrding to the curve.. Or?
Here's my measurements of the subs.. The left and right I have changed to totally flat.. The other on is the one I use with a Harman 4db curve..
I understand When I do volume calibration in Dirac I always set all the speakers the same. My subs (two subs driven by a miniDSP) are always louder than the other speakers so I set them first. But for some reason Dirac still measured them several db lower than the main speakers. I've tried several ways of setting the sub level and the result is always the same. I even set the sub level first and set the other speakers a few db lower ... same result. I sent screen shots to Dirac and exchanged emails with Flavio and never got an answer to that issue. He just said as long as Dirac doesn't throw a clipping error or low signal error the measurements are okay, So I measure in REW after and sometimes I raise or lower the sub level for a smooth transition. And I always use flat targets. Oh.. So this can be a bug in Dirac? Or something with MiniDSP?
I guess something could happen between RMC<>MiniDSP<>subs and having a fourth part calculate the levels.. You don't have the simplest of setups either if I recall correctly..
|
|
|
|
Post by marcl on Dec 5, 2020 12:39:57 GMT -5
I understand When I do volume calibration in Dirac I always set all the speakers the same. My subs (two subs driven by a miniDSP) are always louder than the other speakers so I set them first. But for some reason Dirac still measured them several db lower than the main speakers. I've tried several ways of setting the sub level and the result is always the same. I even set the sub level first and set the other speakers a few db lower ... same result. I sent screen shots to Dirac and exchanged emails with Flavio and never got an answer to that issue. He just said as long as Dirac doesn't throw a clipping error or low signal error the measurements are okay, So I measure in REW after and sometimes I raise or lower the sub level for a smooth transition. And I always use flat targets. Oh.. So this can be a bug in Dirac? Or something with MiniDSP?
I guess something could happen between RMC<>MiniDSP<>subs and having a fourth part calculate the levels.. You don't have the simplest of setups either if I recall correctly..
Yes my setup is not typical for sure! In the end I get what I need, but how Dirac sets levels for measurement and how it translates to the final levels in the XMC-2 remains a mystery. I check the levels after with a SPL meter and always have to tweak channels a couple db. |
|
|
|
Post by megash0n on Dec 5, 2020 14:25:20 GMT -5
Oh.. So this can be a bug in Dirac? Or something with MiniDSP?
I guess something could happen between RMC<>MiniDSP<>subs and having a fourth part calculate the levels.. You don't have the simplest of setups either if I recall correctly..
Yes my setup is not typical for sure! In the end I get what I need, but how Dirac sets levels for measurement and how it translates to the final levels in the XMC-2 remains a mystery. I check the levels after with a SPL meter and always have to tweak channels a couple db. kinda sounds like that same issue I was seeing. Measuring again with the newer versions seemed to fix that. But, I'm convinced you have to use a harmon curve if you want that output to stay high. Otherwise, Dirac is going to normalize it. |
|
|
|
Post by marcl on Dec 6, 2020 7:01:01 GMT -5
|
|
ttocs
Global Moderator 
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 6, 2020 12:49:07 GMT -5
Will there be a quiz later? |
|
|
|
Post by marcl on Dec 6, 2020 17:21:16 GMT -5
Will there be a quiz later? Where are the Snowdons of yesteryear? |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 6, 2020 17:27:30 GMT -5
Will there be a quiz later? Where are the Snowdons of yesteryear? I don't know, I'm a descendent of the Wallace Clan.  |
|
|
|
Post by hsamwel on Dec 11, 2020 17:01:01 GMT -5
Any news of your sub tryouts? Maybe satisfied? |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 11, 2020 17:30:54 GMT -5
Any news of your sub tryouts? Maybe satisfied? Not yet, but I'm very anxious to dive in! Due to a sudden change in schedule of some work in the house, sooner than expected, I won't be having audio fun till tomorrow night. |
|
|
|
Post by BigE on Dec 11, 2020 18:47:42 GMT -5
Any news of your sub tryouts? Maybe satisfied? How do you get your signature line to have items in bold font? |
|
|
|
Post by hsamwel on Dec 12, 2020 6:37:41 GMT -5
Any news of your sub tryouts? Maybe satisfied? How do you get your signature line to have items in bold font? I frame the word with [ B ] and [ /B ] without the spaces. Note that you have limited amount of characters in the sign. When adding these it takes space from the visable as well. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 14, 2020 1:52:35 GMT -5
Got everything going and did some initial calibrations. The Mains and the new subs have Anthem ARC built-in, so I ran ARC on those. Then I got REW involved to tweak the twin sub stacks for Bass Management usage, first as Left Pair and Right Pair, then all four together. There's a Left Stack wired to the Left Sub Output, and the Right Stack is wired to the Right Sub Output. Dirac did ok with each pair, but the four together was not acceptable. Kind of expected that would happen. I was able to tweak to a better result using the controls in the app for these subwoofers from Martin Logan, but since there's another purpose for these subs as dedicated pairs for each Front speaker this will probably need more work. I'd like to keep these two pair connected as they are, two outputs to four subs, but I think it'll end up needing to be all four subs on one output, and, I might need to add a miniDSP to help. Won't know for a few days yet. The two stacks are Martin Logan Dynamo 1100X. They are controllable via an app, and it works very well. Between all the adjustments available and ARC, they're pretty easy to dial-in. The controls in the app are: Volume, Movie/Music modes, LP Filter 30-120Hz (not available when using the LFE input on the sub), Slope, Phase, Invert, 20-30Hz +-10dB. These graphics show which subwoofer does what, along with location.   Still need to work on the Center Sub, and the High Level connection augmented bass for the L&R. One step at a time, so I'll probably work on the Center sub first. It's gonna be a busy day and night so I doubt there will be enough time to post some plots today, probably tomorrow. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 15, 2020 13:32:00 GMT -5
And the FUN begins in earnest!
I now can concentrate on improvements after some involved house work and getting through initial system setups for a Movie Night! last night with a couple friends.
I just took a little time to REW to make the best of what settings are available to get to a baseline without ARC or Dirac. When I ran up against a null which can only be defeated via placement and treatment, which I verified with simple tests of both, I am beginning the small journey of placement, which is something I've been wanting to do for some months now. I figure, whatever makes the twin stacks happy should be good enough for the L&R Mains.
After playing with some MDF boards I've got and some bales of mineral wool and a couple of 2x4 panels, it became apparent that, while the placement of the Right speaker was good for "it", it makes the sub-stack adjacent to it too close to the right wall. I moved the speaker and stack just a few inches and negated the top 5Hz of a 17Hz wide mouth of a deep null which "was" deepest and centered at 69Hz. So if only a few inches of movement plus one 2x4 panel can do that much, then it's time to have more fun.
I did the measurements a long time ago for how far left I want to move the system center, but it will probably move a little more than originally planned - since getting the new subs. Form follows function tempered by lifestyle and comfort. I used to be "all in" with form follows function, but decided I wanted my guests to be happy also.
I've watched 3 movies since Sunday night and all three had one thing in common: Clear Center Channel audio and almost invisible speakers. Knowing that LFE is all by itself in its subwoofer group lets the bass managed audio for the Small speakers provide less distorted sound, with the greatest improvement being the Center Channel. So even the fact that the system was clearly not anywhere near where it should be, the clear dialogue aided in multiplying the enjoyment factor by quite a bit. I never felt like I couldn't understand dialogue. Plus, even though the peak SPL went pretty high it didn't seem overbearing and the clarity of dialogue and sounds during peak moments was still intact!
So the bottom line for now is that Dirac did pretty well with the 7 point run I did on Sunday, not perfect, but pretty well. Dirac didn't get the level balanced right so that'll need to be sorted out on the next run so I don't need to adjust so much. Even so, the timing sounded great so no complaint there.
Since things are a bit wacky I've decided to not post any plots until things are a bit more sorted, especially where Dirac is concerned. Maybe in a day or so.
|
|
|
|
Post by hsamwel on Dec 19, 2020 13:13:23 GMT -5
And the FUN begins in earnest! I now can concentrate on improvements after some involved house work and getting through initial system setups for a Movie Night! last night with a couple friends. I just took a little time to REW to make the best of what settings are available to get to a baseline without ARC or Dirac. When I ran up against a null which can only be defeated via placement and treatment, which I verified with simple tests of both, I am beginning the small journey of placement, which is something I've been wanting to do for some months now. I figure, whatever makes the twin stacks happy should be good enough for the L&R Mains. After playing with some MDF boards I've got and some bales of mineral wool and a couple of 2x4 panels, it became apparent that, while the placement of the Right speaker was good for "it", it makes the sub-stack adjacent to it too close to the right wall. I moved the speaker and stack just a few inches and negated the top 5Hz of a 17Hz wide mouth of a deep null which "was" deepest and centered at 69Hz. So if only a few inches of movement plus one 2x4 panel can do that much, then it's time to have more fun. I did the measurements a long time ago for how far left I want to move the system center, but it will probably move a little more than originally planned - since getting the new subs. Form follows function tempered by lifestyle and comfort. I used to be "all in" with form follows function, but decided I wanted my guests to be happy also. I've watched 3 movies since Sunday night and all three had one thing in common: Clear Center Channel audio and almost invisible speakers. Knowing that LFE is all by itself in its subwoofer group lets the bass managed audio for the Small speakers provide less distorted sound, with the greatest improvement being the Center Channel. So even the fact that the system was clearly not anywhere near where it should be, the clear dialogue aided in multiplying the enjoyment factor by quite a bit. I never felt like I couldn't understand dialogue. Plus, even though the peak SPL went pretty high it didn't seem overbearing and the clarity of dialogue and sounds during peak moments was still intact! So the bottom line for now is that Dirac did pretty well with the 7 point run I did on Sunday, not perfect, but pretty well. Dirac didn't get the level balanced right so that'll need to be sorted out on the next run so I don't need to adjust so much. Even so, the timing sounded great so no complaint there. Since things are a bit wacky I've decided to not post any plots until things are a bit more sorted, especially where Dirac is concerned. Maybe in a day or so. Why would bass mangement+LFE mixed together create distortion? How can one or two subs playing together for each channel play better than let’s say five together calibrated and placed att good positions? What if you use all of the 7 subs for ONE mixed mono channel with a MiniDSP? And then let Dirac calibrate it as one? |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,170
|
Post by ttocs on Dec 19, 2020 16:20:54 GMT -5
1. Why would bass mangement+LFE mixed together create distortion? 2. How can one or two subs playing together for each channel play better than let’s say five together calibrated and placed att good positions? 3. What if you use all of the 7 subs for ONE mixed mono channel with a MiniDSP? And then let Dirac calibrate it as one? 1. LFE is solely for sound effects and low frequency sounds and rumbles. Dolby never intended the LFE audio track to be mixed with any other track. Every channel is supposed to be produced by a "Full Range Speaker. This is not the answer you seek, yet. In a movie with action taking place along with dialogue, the LFE track will be playing heavy, low, thunderous sounds - all of which are below 100Hz - while dialogue is going on. It's reasonable to keep these audio tracks separate so the LFE sounds don't interfere with sound that has more detail in the lower ranges. This can be assumed to be conjecture, until you try it. I separated the LFE into its own subwoofer group using the 3 subs I had at the time. This meant that the bass below the crossovers for all the Small speakers would now be produced by the Large Left/Right speakers with very capable woofers that operate very well up to their 300Hz crossover in the speaker itself. The Center Speaker was instantly clearer and more intelligible!!, and so were the Surrounds. So, my system, in my room, with my subwoofers, having the LFE go away from all the other speakers resulted in a substantial gain in clarity in all the Small speakers. So then I began raising the XO setting for the Center Speaker, which I have had complaint with in regards to the weak low end, and with each raising of the XO there was improvement. At the point where I decided to get more subwoofers, the XO setting was at 150Hz for the Center speaker. Now, I have a subwoofer waiting to be tested, just for the Center channel only. I'll get to that in a few days after working on other things first. 2. This is not really a question with enough clarity to answer with specifics because, in my mind, there are some nuances in what those one or two subs are setup to do. In my case, I have 4 subwoofers in 2 stacks of 2, next to the Left and Right Main speakers. They have two duties: To run as a group of 4 for Bass Management (sorry Keith, I will use BM as a designation for all bass below the XO for all Small speakers) for all Small speakers. And, to run as dedicated pairs for the Left and Right speakers BUT using the High Level Connection (HLC) and used as Augmented Bass for the L&R. My L&R speakers have bass controls for contouring the bass. One is for 75Hz and below, and the other is for Mid-Bass boost and cut. Between these two controls, and ARC, there is quite a lot of help in the bass region before enlisting Dirac for anything. In addition to the above, each subwoofer has some versatile controls plus ARC. The idea of Stack-Of-Subs (SOS) is something promulgated by Rel. I have always liked the idea, but could never find information or demonstrations proving the concept, just a lot of talk by Rel. I "wanted" more subs, so I decided to try this and test it out. I have not had the time do do this yet. I'm working on using the 4 subs for the BM Group. With the work that's been going on within my house, it's something that's been relegated to the little time I spent on it last weekend to get ready for a Movie Night! last Monday. And even though there really was very little time in getting the system up and running after a MAJOR rewiring, the movies we watched sounded the best so far! But, to answer one version of your question directly, they can't. But. Specific intended purpose for the subwoofers comes into play, so it's not so simple for everyone, specifically me. My primary purpose for having the SOS is for the HLC method for augmented bass for the Mains. The secondary usage is for BM. I'm working on each pair to be as good as they can be, and that is working out VERY well! But I also want the 4 subs to play well together, and for this I am getting close to getting another miniDSP. There's a huge dip I will be working on tomorrow, firstly with altering the placement of each SOS and its companion Left or Right speaker but placement options are limited, secondly with room treatment, thirdly by adding a couple doors to my room. I'm in no rush, just want to cover the bases with regard to the foundation of the learning involved. So it's a step by step process. 3. I will never go back to a combined LFE/BM system. I "am" amenable to using more miniDSP's for each multi-sub group so Dirac can calibrate each group as an individual. Again, it's a step by step process. I want to exhaust as much as I am willing and capable of before going on to the next idea. So in the end, the system may (probably) end up being 2 main subwoofer groups for 7 of the 8 subwoofers that are connected to the Center Sub Output for LFE, and Left Sub Output for BM. The HLC for each Left Pair and Right Pair of subs is relatively passive as it is setup to respond from the same speaker connection as the Left and Right speakers, and as such Dirac will calibrate the Left w/SOS and Right w/SOS as full range speakers. The Center Speaker and its Subwoofer will be controlled by a active XO, so the processor will not be involved and Dirac will calibrate as a full range speaker. |
|
|
|
Post by marcl on Dec 20, 2020 10:13:28 GMT -5
ttocs, hsamwel .... I have a little different spin on the idea while being in full agreement on the principle that the LFE channel and bass from small speakers (yes, BM!) are best served by different solutions.
A - I don't believe BM sounds better separated from LFE because the nature of the content causes interference and distortion between them when played through the same speakers. I believe BM sounds better separated from LFE only if the system reproducing the BM is fundamentally different and better suited to the task ... i.e. not just a separate set of subwoofers.
B - I believe LFE requires a solution biased toward much lower frequencies (generally <40Hz, even though LFE spec goes up to 120Hz), needing higher peak output to create the "E" in LFE, and generally not containing the nuances of music or dialogue.
C - I believe BM requires a solution biased toward frequencies between 40Hz and 200Hz, does not need the same peak output as LFE, and has music and dialogue content that benefits from the more articulate reproduction provided by full range speakers. i.e. excellent full range speakers like ttocs' Martin Logans or my Magnepans.
D - As far as I have found in my research, the Harman paper on multiple subs provides the most rigorous mix of simulation and empirical data supporting two or four subs distributed, not with respect to L/R channels, but rather optimized to provide more uniform bass response across listening positions and reducing the magnitude of peaks and nulls ... in acoustically small rooms, in which we all live. REL doesn't provide much science behind their stack idea, though the stack idea does have legitimate origins in large scale sound reinforcement with lots of science describing several solutions ... very complicated physical and phase relationships as well as crossovers and levels. I just don't think the sound reinforcement use case translates to small room acoustics. They are never dealing with modal resonances.
So ... again agreeing in principle, I like a solution that definitely sets LFE apart with its own speakers and amps. I believe BM is best served sent to full range L/R speakers, perhaps augmented with subs only <40Hz. This is how my system is set up now.
p.s. I really think this is a great topic and I honestly have never seen it discussed anywhere else! Kind of amazing.
|
|
