|
|
Post by igorzep on Sept 2, 2014 9:18:04 GMT -5
Basically the way the XMC-1 should handle dual subs is the same Way The processors with Audyysey XT32 do. Audyysey XT32 with the dual sub management is absolutely incredible in the job it does. What I'm expecting from my xmc-1 once it arrives is to do a better job with room correction, but if it doesn't handle sual subs properly, thats not going to be possible. Why should it copy the worst part of the Audyssey calibration? How can the job done no better, or even worse than Y connector would do is incredible? XT32 does well, but SubEQ HT is a joke, what it does is counter productive to optimize sub interaction. |
|
|
|
Post by AudioHTIT on Sept 2, 2014 9:42:02 GMT -5
^^^ Somehow you attributed a comment to urwi that was made by smfins (and your comment by smfins). |
|
|
|
Post by igorzep on Sept 2, 2014 9:47:33 GMT -5
I'll ask bootman, igorzep and anyone else to chime in if I missed something. Here's what we need an answer to: 1. In dual mono mode, will Dirac issue unique test tones to each subwoofer for the purpose of setting distance and level independently for each sub? 2. In dual mono mode, will Dirac issue a simultaneous test tone from both subs for the sake of EQ/room correction? The answer is obvious 1) If it is presented with two subs Dirac will measure level/distance and calculate EQ curve for each one, in isolation, as with any other channel. 2) If it is presented with two subs - they are just two channels for Dirac, it will do, as the above - calibrate one by one. It can be said that it doesn't even know those channels are Subwoofers - everything is just a channel for DL. The solution of the problem I've posted long ago. The idea is as simple as it can be: 1) Allow to do relative phase/level/delay settings for subs manually. 2) Present both as one channel to the DL engine respecting/keeping the relative manual settings. The measured data from DL will still be useful for time-aligning the 'combined' channel with satellites. Then what we have: It is not quite trivial to setup optimally (but so is the Dual Sub setup by itself) No magic tricks Easy to implement Is very flexible And contrary to 'convenient' solutions - it will work, not just do 'magic' that is not working! Otherwise we have two sub channels that are impossible to optimize better than just an Y-splitter would do. So we have a 7.1 working unit and a .1 bit of marketing. |
|
|
|
Post by igorzep on Sept 2, 2014 9:50:40 GMT -5
^^^ Somehow you attributed a comment to urwi that was made by smfins (and your comment by smfins). Hopefully corrected it as it was intended  Quoting is not so convenient here  |
|
|
|
Post by bluescale on Sept 2, 2014 12:50:42 GMT -5
Yes indeed I've misread your post. Sorry for that. I don't see how Dirac Live could work with dual mono subs. They probably would need to make changes to their software to suport that case, i.e. set level and delay for each sub then EQ as one unit. For the record, despite this limitation with Dirac, Mark Seaton said that Dirac did a phenomenal job with 4 subs in an installation he did recently. I'm still trying to wrap my head around how it worked, but I trust that man's experience over my limited knowledge. |
|
|
|
Post by bluescale on Sept 2, 2014 12:58:30 GMT -5
1) If it is presented with two subs Dirac will measure level/distance and calculate EQ curve for each one, in isolation, as with any other channel. 2) If it is presented with two subs - they are just two channels for Dirac, it will do, as the above - calibrate one by one. It can be said that it doesn't even know those channels are Subwoofers - everything is just a channel for DL. The solution of the problem I've posted long ago. The idea is as simple as it can be: 1) Allow to do relative phase/level/delay settings for subs manually. 2) Present both as one channel to the DL engine respecting/keeping the relative manual settings. The measured data from DL will still be useful for time-aligning the 'combined' channel with satellites. What do you think of the possible workaround I posted earlier. Any reason this wouldn't work for a *mostly* automated system: 1. Set mode to Stereo subwoofer 2. Have Dirac complete the distance and level calibration 3. Set mode to Dual Mono subwoofer 4. Have Dirac complete the EQ The only caveat I see here is whether a subwoofer mode change is allowed in the middle of a Dirac calibration. Is there anything else I'm missing? |
|
|
|
Post by igorzep on Sept 2, 2014 13:21:45 GMT -5
Any reason this wouldn't work for a *mostly* automated system: 1. Set mode to Stereo subwoofer 2. Have Dirac complete the distance and level calibration 3. Set mode to Dual Mono subwoofer 4. Have Dirac complete the EQ The only caveat I see here is whether a subwoofer mode change is allowed in the middle of a Dirac calibration. Is there anything else I'm missing? 1) There is no "mode" inside XMC-1. There is a digital signal chain with digital filters controlled programmatically. A 'Mode' is a user level concept. There is no point to play magic with user modes if you have control on internal state anyway... 2) Dirac works as complete process: Take measurements -> return measured data together with filters for every channel back to AVR. It is not that something can be inserted inside the process. 3) It will be as incorrect as the SubEQ HT... what is the point... The distance from the sub to the main listening position in not the right value for the delay. It is right only in one case - when both subs are at the front wall placed symmetrically and main measurement position is perfectly centered. But then... the result is equal with Y connector and there is no need for second sub out. |
|
bootman
Emo VIPs 
Typing useless posts on internet forums....
Posts: 9,358
|
Post by bootman on Sept 2, 2014 13:26:30 GMT -5
We are going around in circles.
We need someone from Emotiva to definitively tell us how this is going to work with multiple subs.
I want to know how it works in "Stereo mode" and in "Dual mono mode".
Then we can make an informed decision if the unit meets one's need with regards to multiple subs.
|
|
|
|
Post by igorzep on Sept 2, 2014 13:58:43 GMT -5
^
I think we got the answer quite clear. Dual mono mode == Stereo mode with bass joined together before it sent to the sub delay/level/EQ.
I am getting in only because there are again repeated suggestions to do something that have no value. Someone already were suggesting repeatedly independent EQ... so - we have now what we have - with Dirac there is one mode 'convenient' but useless "bass management" mode.
|
|
KeithL
Administrator 
Posts: 10,273
|
Post by KeithL on Sept 2, 2014 16:48:23 GMT -5
Dirac treats each speaker as separate for purposes of calibration - and that DOES include the subs. Since Dirac's measurements include impulse response and distance, the calibration process will correct for "phase differences" between the subs - whether they are caused by differences in distance or other differences between the subs (or even by different settings on the subs themselves). (So, yes, the subs are independently EQed). If you then select Dual Mono Subs, the LFE information, and the bass below the crossovers for all of your other speakers, will be summed and sent to both subs. (Again, if the subs are different distances from the listening position, their distances will be "compensated for" just as the distances from the listening position to each of the other speakers.) If you select Stereo Subs, the LFE channel will be sent to BOTH subs, the bass (below the crossover points) from all your left channel speakers will go to the Left Sub, the bass (below the crossover points) from all your right channel speakers will go to the Right Sub, and the bass (below the crossover points) from the center channel will go to BOTH subs. We are going around in circles. We need someone from Emotiva to definitively tell us how this is going to work with multiple subs. I want to know how it works in "Stereo mode" and in "Dual mono mode". Then we can make an informed decision if the unit meets one's need with regards to multiple subs. |
|
|
|
Post by bluescale on Sept 3, 2014 0:00:47 GMT -5
I just went back to the Dirac manual, because it's been almost two months since I used the trial. I found this little gem I hadn't seen before:
Is this option going to be available in DiracLive LE?
|
|
|
|
Post by bluescale on Sept 3, 2014 0:05:13 GMT -5
Since Dirac's measurements include impulse response and distance, the calibration process will correct for "phase differences" between the subs - whether they are caused by differences in distance or other differences between the subs (or even by different settings on the subs themselves). How is that possible? That's what we don't understand... |
|
|
|
Post by bluescale on Sept 3, 2014 15:23:58 GMT -5
Okay, a little change from our Dirac questions. I did some measurements of various modes, while trying to look into weirdness in All Stereo mode. When I look at modes OTHER THAN All Stereo, the bass management looks right:  However, Roger Dressler pointed out this anomaly to me: I believe Chuck Elliot may have been trying to ask me a similar question this morning, and I just didn't understand his question. KeithL - I'd love to hear your thoughts on this. I'm still learning how this stuff should works and should look, so I'm a little unclear. Obviously, I trust Roger's knowledge on on the subject implicitly. Did I do something wrong in my testing, or is something misbehaving here? For the record, all speakers were set to small and xovered at 80hZ. The low pass filter slope was set to 24dB. All settings for EQ, distance and level were set to 0. |
|
|
|
Post by Chuck Elliot on Sept 3, 2014 15:54:06 GMT -5
Just FYI, if these are suppose to be Linkwitz–Riley filter slopes they should be 6 dB down at the crossover point. This allows them to sum flat with equal slope. I find it too bad that the high-pass is fixed as there are times when both being at 24 might be desirable. A 12 dB LR high-pass and 24 dB LR low-pass should look like this:  With both at 24 dB:  Perhaps someone at Emotiva can explain more.... |
|
|
|
Post by igorzep on Sept 3, 2014 16:09:49 GMT -5
I find it too bad that the high-pass is fixed as there are times when both being at 24 might be desirable. Is high-pass fixed in XMC-1? While low-pass is selectable between 12/24 dB/octave? Shouldn't it be vice-versa... A 12 dB LR high-pass and 24 dB LR low-pass should look like this: 24 dB LP - correct, but shouldn't 12 dB HP be -3dB at crossover point (and after summing with natural speaker 12 dB/octave rolloff at the same frequency this will become -6dB in total and LR4 acoustically from both sides)? |
|
|
|
Post by Chuck Elliot on Sept 3, 2014 16:18:57 GMT -5
I find it too bad that the high-pass is fixed as there are times when both being at 24 might be desirable. High-pass if fixed in XMC-1? While low-pass is selectable between 12/24 dB/octave? Shouldn't it be vice-versa... A 12 dB LR high-pass and 24 dB LR low-pass should look like this: 24 dB LP - correct, but shouldn't 12 dB HP be -3dB at crossover point (and after summing with natural speaker 12 dB/octave rolloff at the same frequency this will become -6dB in total and LR4 acoustically from both sides)? LR filters are 6 dB down at the crossover point by definition! Butterworth and other filters are -3 dB. A 12 dB hi-pass IS desirable for some systems. This is particularly true when the natural roll-off is close to the crossover frequency in a sealed enclosure. 12 dB elect + 12 dB roll-off = 24 dB. |
|
|
|
Post by igorzep on Sept 3, 2014 16:24:13 GMT -5
LR filters are 6 dB down at the crossover point by definition! Butterworth and other filters are -3 dB. A 12 dB hi-pass IS desirable for some systems. This is particularly true when the natural roll-off is close to the crossover frequency in a sealed enclosure. 12 dB elect + 12 dB roll-off = 24 dB. It is exactly what I am talking about. The acoustically they both should be LR4 and -6dB at the crossover point, again - acoustically. But High-Pass LR4 is formed from 2 second-order roll-offs - electrical and acoustical, and so each one of them should have -3dB at this point so it will sum to required -6dB acoustically. |
|
|
|
Post by bluescale on Sept 3, 2014 16:26:08 GMT -5
Is high-pass fixed in XMC-1? While low-pass is selectable between 12/24 dB/octave? Shouldn't it be vice-versa...[/quote] From the manual: Isn't that how most processors do it? |
|
|
|
Post by igorzep on Sept 3, 2014 16:34:44 GMT -5
From the manual: Isn't that how most processors do it? 12 dB/octave is useless for Low Pass. But for High Pass both 12 and 24 dB/octave are useful, for the reason Chuck Elliot explained: 1) 12 dB/octave for case when we are crossing near the natural speaker rolloff 2) 24 dB/octave if we are crossing significantly above the natural rolloff of the speaker Both cases target 24 dB/octave acoustically. Can't say for other processors for sure, or even most of them, but this is what is logical to do |
|
|
|
Post by Chuck Elliot on Sept 3, 2014 16:46:11 GMT -5
LR filters are 6 dB down at the crossover point by definition! Butterworth and other filters are -3 dB. A 12 dB hi-pass IS desirable for some systems. This is particularly true when the natural roll-off is close to the crossover frequency in a sealed enclosure. 12 dB elect + 12 dB roll-off = 24 dB. It is exactly what I am talking about. The acoustically they both should be LR4 and -6dB at the crossover point, again - acoustically. But High-Pass LR4 is formed from 2 second-order roll-offs - electrical and acoustical, and so each one of them should have -3dB at this point so it will sum to required -6dB acoustically. I hear what you're trying to say, but it doesn't work that way. The following graph includes 2 new traces: Yellow is the response of the loudspeaker with natural roll-off. F3 is about 58Hz. Orange shows the sum of the roll-off and electrical crossover. The net effect is that just the slope increases. The -6 dB point moves very little. This is from my own design software as I've spent a bunch of time analyzing bass-management with it.  |
|
Quoting is not so convenient here 
