|
|
Post by jagman on Jan 26, 2021 12:02:23 GMT -5
I have a recently acquired RMC1 and have been dabbling with DIRAC. So far the results have sounded quite good except for dealing with the Subwoofer.
I have a 9.1.6 alignment and 2 rows of seats. The MLP is the center seat in the second row. The closest speakers there are the sides, rear surrounds followed by top rear. The top middle are by far the least efficient speakers because they are in an IB alignment vs in boxes. The LCR are the same speaker. The rest are all the same speaker (with the exception of the top middle speakers not having an enclosure).
When I take REW sweeps of the L and SW pre DIRAC and then post DIRAC, the SW lose about 15dB output with DIRAC enabled. I presume that is because they have been level matched to the top middle speakers. The interesting thing is instead of taming the peaks in the SW response to bring it down match the top middle speakers, DIRAC is merely lowering the curve of the SW without altering it's shape, so that means I'm giving up the ULF. If I want to put in a house curve and raise the SW output well above the post DIRAC curve and add back in the ULF, it lets me. Doing a REW sweeps after this still has output less than the original unprocessed curve, but a better shape. Is this DIRAC understanding what the SW was capable of before it neutered it's response thus I don't need to be worried about over boosting the bass?
|
|
|
|
Post by marcl on Jan 26, 2021 12:02:28 GMT -5
cwmcobra rhale64 Thank you! Trying to keep myself out of trouble now that I retired.  |
|
|
|
Post by marcl on Jan 26, 2021 12:16:35 GMT -5
I have a recently acquired RMC1 and have been dabbling with DIRAC. So far the results have sounded quite good except for dealing with the Subwoofer. I have a 9.1.6 alignment and 2 rows of seats. The MLP is the center seat in the second row. The closest speakers there are the sides, rear surrounds followed by top rear. The top middle are by far the least efficient speakers because they are in an IB alignment vs in boxes. The LCR are the same speaker. The rest are all the same speaker (with the exception of the top middle speakers not having an enclosure). When I take REW sweeps of the L and SW pre DIRAC and then post DIRAC, the SW lose about 15dB output with DIRAC enabled. I presume that is because they have been level matched to the top middle speakers. The interesting thing is instead of taming the peaks in the SW response to bring it down match the top middle speakers, DIRAC is merely lowering the curve of the SW without altering it's shape, so that means I'm giving up the ULF. If I want to put in a house curve and raise the SW output well above the post DIRAC curve and add back in the ULF, it lets me. Doing a REW sweeps after this still has output less than the original unprocessed curve, but a better shape. Is this DIRAC understanding what the SW was capable of before it neutered it's response thus I don't need to be worried about over boosting the bass? Not sure exactly what's going on, but it could be related to the actual measurement levels of the sub vs other speakers. Could you do a screen shot like this ...  This shows my Center Sub highlighted. You can faintly see that the 35Hz peak on my L/R Fronts (large) and the 35Hz peak in the Sub measurement are at about the same level. Once I got Dirac to do this properly, the corrections and listening levels of all the speakers worked fine. Doing filters on a measurement like this (7.1.4, Center Sub=LFE) with flat target curves, results in REW 7.1 measurement like this ...  This is pretty good. The reason the small speakers have elevated bass is because their bass is directed to the large L/R speakers. This is an issue that has been reported and hopefully will be fixed. I did have to tweak levels a bit in the Preset to get them within .5db. |
|
|
|
Post by phantom3rdchannel on Jan 26, 2021 17:21:03 GMT -5
My room is very well treated and my imaging and sound stage without Dirac are pristine. Reach out and touch the singer kind of imaging. With Dirac, its a mess. It's either Left or Right or indistinguishable (which sounds exactly like an out of phase issue). My Mic is the EMO supplied mic on a stand. The primary listening position is exactly the same distance from the left and right speaker, the room is 100% symmetrical. The speakers are exactly the same distance from all room boundaries (Side walls, back walls front walls, ceiling.) The walls are all treated the same and the speakers measure almost identical before any filters are applied. I measured every mic position, and made sure that the left and right measurements were in the same relative positions. I have done this multiple times. I widened and narrowed the mic positioning. I raised the mics, I lowered the mics. Results were sometimes different but imaging was always smeared or off center by a lot. I have not even come close to getting a good center image. Its funny, before I bought my XMC-2, I had a Pioneer Elite AV Receiver. It was second to the top of the line in 2005. It had Pioneer Room Correction software built in. It took 20 minutes to set it up and run it. The results were amazing, The main reason I replaced it was because is was pre HDMI and switching between sources for Home Theater using external HDMI switches became too complex and unreliable. So I upgraded to EMO XMC2 and an EMO Amp. I am mostly impressed by the Emo sound quality, but the Dirac software is just unacceptable. I have reverted to REW and the internal PEQs. In about an hour I had achieved very listenable results with a dead on center image and a wide sound stage. If you are suffering through the Dirac issues, like I have been, I highly recommend that you learn how to use the internal PEQs. For 2.0 or 2.1 Music its pretty easy. For some reason Dirac seems to have worked in my room for 7.1 movies so I am still using that for Movies and TV. I may try Dirac again in about a year to see if they fix the 2 channel image problem, but for now, I don't need it. I feel free!!!! Have you looked at the impulse response of your left and right speakers in REW? I have an issue that's caused by some reflections due to my L/R speaker setup (i.e. not the fault of Dirac or the processor) and I have a way to mitigate this issue using a temporary absorber when I do the MLP measurement with Dirac. That part isn't relevant to your issue, but the symptom might be similar. When Dirac is fooled by my circumstance I get impulse response like this ... a small spike on the right channel which Dirac incorrectly uses to align L/R, resulting in the right being delayed 1.5ms. I can't say for sure this causes an imaging issue, but if it did it would skew the image a little to the left. Do you see anything like this, or are your L/R impulses perfectly aligned at 0.0ms? View AttachmentThanks, I haven't looked at this, I will try it next time I get a chance to take some measurements. |
|
|
|
Post by markc on Jan 28, 2021 1:37:10 GMT -5
Do you set up the curtains for what your speakers/sub frequency can handle or just leave the dirac stock settings ? The screenshots you sent are not at all helpful I'm afraid. On first glance I thought you had an impossibly great system but then I realised you only posted the predicted "corrected curves" We would need to see the actual measured traces for each speaker to help you by commenting. Check the "measured" box to display the actual measured response, overlayed with your target and then we can see how much you are pushing the system with filters to get the predicted. Without knowing how much you are using Dirac's filters to try and get close to those "ideal result" curves you've drawn, we have no way to usefully comment about your target curves or curtains and whether they are optimal or realistic. |
|
|
|
Post by jagman on Jan 28, 2021 1:38:54 GMT -5
I have a recently acquired RMC1 and have been dabbling with DIRAC. So far the results have sounded quite good except for dealing with the Subwoofer. I have a 9.1.6 alignment and 2 rows of seats. The MLP is the center seat in the second row. The closest speakers there are the sides, rear surrounds followed by top rear. The top middle are by far the least efficient speakers because they are in an IB alignment vs in boxes. The LCR are the same speaker. The rest are all the same speaker (with the exception of the top middle speakers not having an enclosure). When I take REW sweeps of the L and SW pre DIRAC and then post DIRAC, the SW lose about 15dB output with DIRAC enabled. I presume that is because they have been level matched to the top middle speakers. The interesting thing is instead of taming the peaks in the SW response to bring it down match the top middle speakers, DIRAC is merely lowering the curve of the SW without altering it's shape, so that means I'm giving up the ULF. If I want to put in a house curve and raise the SW output well above the post DIRAC curve and add back in the ULF, it lets me. Doing a REW sweeps after this still has output less than the original unprocessed curve, but a better shape. Is this DIRAC understanding what the SW was capable of before it neutered it's response thus I don't need to be worried about over boosting the bass? Not sure exactly what's going on, but it could be related to the actual measurement levels of the sub vs other speakers. Could you do a screen shot like this ... View AttachmentThis shows my Center Sub highlighted. You can faintly see that the 35Hz peak on my L/R Fronts (large) and the 35Hz peak in the Sub measurement are at about the same level. Once I got Dirac to do this properly, the corrections and listening levels of all the speakers worked fine. Doing filters on a measurement like this (7.1.4, Center Sub=LFE) with flat target curves, results in REW 7.1 measurement like this ... View AttachmentThis is pretty good. The reason the small speakers have elevated bass is because their bass is directed to the large L/R speakers. This is an issue that has been reported and hopefully will be fixed. I did have to tweak levels a bit in the Preset to get them within .5db. Here are two images. This first is the raw measurements (which I saved on my PC)  The second is after DIRAC filters were applied.  You can see the DIRAC filters have really brought down the subwoofer output. Going in I expected the DIRAC filters to have tamed the peaks so the valleys would no longer be present, but that didn't happen. The valleys in the sub output are still there, just at a lower dB. Edit: when level setting the speakers to 15dB in DIRAC, I had to set the subwoofer to -8dB. I tried not bringing it down but DIRAC said it was clipping. |
|
|
|
Post by motogp34 on Jan 28, 2021 2:13:06 GMT -5
Do you set up the curtains for what your speakers/sub frequency can handle or just leave the dirac stock settings ? The screenshots you sent are not at all helpful I'm afraid. On first glance I thought you had an impossibly great system but then I realised you only posted the predicted "corrected curves" We would need to see the actual measured traces for each speaker to help you by commenting. Check the "measured" box to display the actual measured response, overlayed with your target and then we can see how much you are pushing the system with filters to get the predicted. Without knowing how much you are using Dirac's filters to try and get close to those "ideal result" curves you've drawn, we have no way to usefully comment about your target curves or curtains and whether they are optimal or realistic. |
|
|
|
Post by motogp34 on Jan 28, 2021 2:13:36 GMT -5
Do you set up the curtains for what your speakers/sub frequency can handle or just leave the dirac stock settings ? The screenshots you sent are not at all helpful I'm afraid. On first glance I thought you had an impossibly great system but then I realised you only posted the predicted "corrected curves" We would need to see the actual measured traces for each speaker to help you by commenting. Check the "measured" box to display the actual measured response, overlayed with your target and then we can see how much you are pushing the system with filters to get the predicted. Without knowing how much you are using Dirac's filters to try and get close to those "ideal result" curves you've drawn, we have no way to usefully comment about your target curves or curtains and whether they are optimal or realistic. |
|
|
|
Post by motogp34 on Jan 28, 2021 2:13:56 GMT -5
Do you set up the curtains for what your speakers/sub frequency can handle or just leave the dirac stock settings ? The screenshots you sent are not at all helpful I'm afraid. On first glance I thought you had an impossibly great system but then I realised you only posted the predicted "corrected curves" We would need to see the actual measured traces for each speaker to help you by commenting. Check the "measured" box to display the actual measured response, overlayed with your target and then we can see how much you are pushing the system with filters to get the predicted. Without knowing how much you are using Dirac's filters to try and get close to those "ideal result" curves you've drawn, we have no way to usefully comment about your target curves or curtains and whether they are optimal or realistic. Attachments:
|
|
|
|
Post by markc on Jan 28, 2021 4:17:30 GMT -5
motogp34Thinking about your question - I think one should set the curtains according to what the speakers can realistically be asked to achieve. I assume that the Dirac's "automatically suggested" curtains are doing the same thing, using an algorithm to suggest curtains. You do have lovely flat speaker-room response measurements even before Dirac is asked to go to work! Just some minor variations across the frequency spectrum except for... The subwoofer! Good low range performance but really dropping off from 100Hz. Are you using a built in or external Low Pass Filter on the subwoofer??? Otherwise I can't believe that a sub with such good response to 20Hz is that poor with the higher frequencies it is meant to deliver. By specification, the .1 LFE channel contains audio up to 150Hz and using a crossover to divert bass for "small" speakers would require good sub performance for at least 1/2 and octave above the crossover - i.e. for an 80Hz crossover, the subwoofer would need to be at least reasonably flat until 120Hz Finally, sorry to be a pain but the Dirac software obscures all of the measurements outside the set curtains, so we still can't comment to answer your original question of whether Dirac's suggested curtains could be tweaked or are already optimal. We would need to see the speaker/room's natural roll off capability
|
|
|
|
Post by marcl on Jan 28, 2021 4:30:36 GMT -5
The screenshots you sent are not at all helpful I'm afraid. On first glance I thought you had an impossibly great system but then I realised you only posted the predicted "corrected curves" We would need to see the actual measured traces for each speaker to help you by commenting. Check the "measured" box to display the actual measured response, overlayed with your target and then we can see how much you are pushing the system with filters to get the predicted. Without knowing how much you are using Dirac's filters to try and get close to those "ideal result" curves you've drawn, we have no way to usefully comment about your target curves or curtains and whether they are optimal or realistic. I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. |
|
|
|
Post by motogp34 on Jan 28, 2021 11:20:51 GMT -5
I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. Are the white lines Dirac stock set curtains ? |
|
|
|
Post by motogp34 on Jan 28, 2021 11:41:32 GMT -5
motogp34 Thinking about your question - I think one should set the curtains according to what the speakers can realistically be asked to achieve. I assume that the Dirac's "automatically suggested" curtains are doing the same thing, using an algorithm to suggest curtains. You do have lovely flat speaker-room response measurements even before Dirac is asked to go to work! Just some minor variations across the frequency spectrum except for... The subwoofer! Good low range performance but really dropping off from 100Hz. Are you using a built in or external Low Pass Filter on the subwoofer??? Otherwise I can't believe that a sub with such good response to 20Hz is that poor with the higher frequencies it is meant to deliver. By specification, the .1 LFE channel contains audio up to 150Hz and using a crossover to divert bass for "small" speakers would require good sub performance for at least 1/2 and octave above the crossover - i.e. for an 80Hz crossover, the subwoofer would need to be at least reasonably flat until 120Hz Finally, sorry to be a pain but the Dirac software obscures all of the measurements outside the set curtains, so we still can't comment to answer your original question of whether Dirac's suggested curtains could be tweaked or are already optimal. We would need to see the speaker/room's natural roll off capability. I set the curtains according to the specs of the speaker ? Here is the link to my sub, seatonsound.net/product/submersivehp/ Not using a external low pass filter. sub is set too +/-3dB from 19Hz to 200Hz outdoor response. Do you recommend all speakers set to small ? |
|
|
|
Post by marcl on Jan 28, 2021 12:18:32 GMT -5
I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. Are the white lines Dirac stock set curtains ? The white dashed lines are Dirac's recommendation of what the speaker's useable response is ... as measured in your room. I think it's pretty accurate. They're just a reference. The curtains are movable. to the right of the right curtain, and to the left of the left curtain, Dirac does nothing. You can move the curtains, but what Dirac does between the curtains is determined by the target curve. That's why I say you can move the curtain wider in order to force Dirac to limit response as well as extend. Example of how I limit my sub response range with curtains and control points:  |
|
|
|
Post by hsamwel on Jan 28, 2021 14:14:21 GMT -5
motogp34 Thinking about your question - I think one should set the curtains according to what the speakers can realistically be asked to achieve. I assume that the Dirac's "automatically suggested" curtains are doing the same thing, using an algorithm to suggest curtains. You do have lovely flat speaker-room response measurements even before Dirac is asked to go to work! Just some minor variations across the frequency spectrum except for... The subwoofer! Good low range performance but really dropping off from 100Hz. Are you using a built in or external Low Pass Filter on the subwoofer??? Otherwise I can't believe that a sub with such good response to 20Hz is that poor with the higher frequencies it is meant to deliver. By specification, the .1 LFE channel contains audio up to 150Hz and using a crossover to divert bass for "small" speakers would require good sub performance for at least 1/2 and octave above the crossover - i.e. for an 80Hz crossover, the subwoofer would need to be at least reasonably flat until 120Hz Finally, sorry to be a pain but the Dirac software obscures all of the measurements outside the set curtains, so we still can't comment to answer your original question of whether Dirac's suggested curtains could be tweaked or are already optimal. We would need to see the speaker/room's natural roll off capability. I set the curtains according to the specs of the speaker ? Here is the link to my sub, seatonsound.net/product/submersivehp/ Not using a external low pass filter. sub is set too +/-3dB from 19Hz to 200Hz outdoor response. Do you recommend all speakers set to small ? I usually follow the room response. My subs are speced from +/- 3db 19-180hz, similar to yours. Here are a couple of my speakers, Center sub (dual subs), Center and Right Front.
On the center I have smooth out the drop before 47hz with added curve points. The Right Front did not need this as it drops nicely. I use the curtain to follow the drop so I don't exceed the speaker response.
The sub I have smooth out the high freqency drop. Got the nasty drop at about 120hz after I added PEQ in the subs to fix a couple of dips at 35hz and 65hz.
|
|
|
|
Post by leonski on Jan 28, 2021 18:54:24 GMT -5
I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. Can you name ANY music with 16hz included? I can think of ONE piece and it needs a very special pipe organ, since most do not include a 64' pipe...... And since nobody knows what movie effects are Supposed to sound like, that part of the discussion is moot. |
|
|
|
Post by marcl on Jan 28, 2021 19:20:21 GMT -5
I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. Can you name ANY music with 16hz included? I can think of ONE piece and it needs a very special pipe organ, since most do not include a 64' pipe...... And since nobody knows what movie effects are Supposed to sound like, that part of the discussion is moot. Saint Saens Symphony #3, the "Organ Symphony" has the 32' pedal rank and I have measured the 16Hz in my room. I have a couple other pipe organ recordings that have the 32', including this one ... The Aeolian Organ at Duke University Chapel - Pentatone  The Bosendorfer Imperial Grand Piano goes down to 16Hz. Of course synthesizers can do what they will. The Dolby spec for LFE probably allows for sounds south of 20Hz but I bet it's rare. No, I don't think anyone would say you need to go below 20Hz. And there's good reason to roll off so there are no subsonics wasting energy. Likewise, I roll my subs off above 120Hz because LFE only goes up that high (theoretically) and no need to reproduce any distortion. |
|
|
|
Post by megash0n on Jan 28, 2021 19:53:59 GMT -5
I (mostly) agree with markc. I set my curtains just outside the white dashed vertical lines, but not much. I also use flat target curves ... but that's just me, apparently. One caution ... I would not pull the subwoofer response down that low! There is absolutely no content from any source below 16Hz. And in fact, I pull my left curtain a little wide to the left and then add a couple control points to intentionally put a steep rolloff below 16Hz. But your speakers look great and should really sound amazing once optimized with Dirac. Can you name ANY music with 16hz included? I can think of ONE piece and it needs a very special pipe organ, since most do not include a 64' pipe...... And since nobody knows what movie effects are Supposed to sound like, that part of the discussion is moot. what exactly would "sound" have to do with it? Are you referring to sound pressure or actually hearing it? Aren't both affected by the same variables? |
|
ttocs
Global Moderator 
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,125
|
Post by ttocs on Jan 28, 2021 21:32:30 GMT -5
Can you name ANY music with 16hz included? I can think of ONE piece and it needs a very special pipe organ, since most do not include a 64' pipe...... And since nobody knows what movie effects are Supposed to sound like, that part of the discussion is moot. Saint Saens Symphony #3, the "Organ Symphony" has the 32' pedal rank and I have measured the 16Hz in my room. I have a couple other pipe organ recordings that have the 32', including this one ... The Aeolian Organ at Duke University Chapel - Pentatone View AttachmentThe Bosendorfer Imperial Grand Piano goes down to 16Hz. Of course synthesizers can do what they will. The Dolby spec for LFE probably allows for sounds south of 20Hz but I bet it's rare.No, I don't think anyone would say you need to go below 20Hz. And there's good reason to roll off so there are no subsonics wasting energy. Likewise, I roll my subs off above 120Hz because LFE only goes up that high (theoretically) and no need to reproduce any distortion. Even though I do like you do with the Low Dirac Curtain - pull it down and add Control Points to cutoff the bass below 18Hz, my Fronts seem to measure with more output below 18Hz than I expected. Red is both Fronts playing mono together.  The Dolby Spec for LFE, otherwise known as the Screen Subwoofer: " 3 Screen Subwoofer
3.1 Sound Pressure Level: +10 dB (Compared to Center Loudspeaker)
The Low-Frequency Effects channel subwoofer must have a flat response over the range of 31.5 to 120 Hz. When compared with a full-range screen channel, the subwoofer channel must be capable of producing +10 dB of in-band gain (for example, as viewed on a real-time analyzer)." The Full Range Screen Speakers (LCR) have the same spec at the low end of the frequency range, whereas, the Surround Speakers are spec'd at 40Hz at the low end which includes the overheads. The above doesn't stop movies from having lower frequencies than spec'd. I've measured strong low 20Hz range in the Center Channel, and lower than 20Hz in the LFE Channel. I'm gearing up for some measurement fun this weekend to check some theories running through my brain about what might be some issues between system setup with Center Sub LFE vs Center Sub Mono. "I gots to know". (Dirty Harry) |
|
|
|
Post by jagman on Jan 28, 2021 23:26:17 GMT -5
Not sure exactly what's going on, but it could be related to the actual measurement levels of the sub vs other speakers. Could you do a screen shot like this ... View AttachmentThis shows my Center Sub highlighted. You can faintly see that the 35Hz peak on my L/R Fronts (large) and the 35Hz peak in the Sub measurement are at about the same level. Once I got Dirac to do this properly, the corrections and listening levels of all the speakers worked fine. Doing filters on a measurement like this (7.1.4, Center Sub=LFE) with flat target curves, results in REW 7.1 measurement like this ... View AttachmentThis is pretty good. The reason the small speakers have elevated bass is because their bass is directed to the large L/R speakers. This is an issue that has been reported and hopefully will be fixed. I did have to tweak levels a bit in the Preset to get them within .5db. Here are two images. This first is the raw measurements (which I saved on my PC) The second is after DIRAC filters were applied. You can see the DIRAC filters have really brought down the subwoofer output. Going in I expected the DIRAC filters to have tamed the peaks so the valleys would no longer be present, but that didn't happen. The valleys in the sub output are still there, just at a lower dB. Edit: when level setting the speakers to 15dB in DIRAC, I had to set the subwoofer to -8dB. I tried not bringing it down but DIRAC said it was clipping. bump. BTW, the bass definitely sounds neutered. It's cleaner and the midrange and imaging is significantly improved but movies lose their impact and it's quite noticeable. |
|
