|
Post by ansat on Apr 26, 2015 14:18:04 GMT -5
EDIT :: Made a correction to the chart labels. Nick -- I did some extensive testing this morning and here is what I found. All tests are measured from the XMC-1 XLR outputs. (no speakers, no microphone)
1. The crossover ignores the gain structure set by dirac for the sub and appears to apply the gain structure of the mains to the sub instead. 2. Filters from the mains and the sub are applied to audio sent to the sub from the crossover. 2a. You can stop this by setting the curtain of the speaker that is being crossed over to the crossover point. (this will only allow the subs filter to affect the response)
NOTE -- These graphs do not measure the same way that speaker measurements would. There is no room gain and no speakers involved. The higher gain of the LFE measurement would be normal since it requires more gain to get the subwoofer to play the same SPL as the mains. (My Subs are about 6db less efficient then the mains) Also, the crossovers would not match correctly due to this. However, I did find that the LFE is about 10db hot compared to when they play from the LFE then played from the xover. Not sure why this is like this and I don't have another branded receiver to test with. Below is the graphs of the original filter. (no custom curtains applied)
Blue = INPUT - LFE OUTPUT - SUBWOOFER Red = INPUT - FRONT OUTPUT - SUBWOOFER - 80hz X-Over Purple = INPUT - FRONT OUTPUT - FRONT - FULL RANGE Yellow = INPUT - FRONT OUTPUT - FRONT - 80hz X-Over
Below is the graphs of new filter with curtain set at the crossover point. (curtains were not changed for the subwoofer.)
Purple = INPUT - LFE OUTPUT - SUBWOOFER Green = INPUT - FRONT OUTPUT - SUBWOOFER - 80hz X-Over Blue = INPUT - FRONT OUTPUT - FRONT - 80hz X-Over
REW file
Tony
|
|
tubby
Emo VIPs
Route 2 in Weekapaug!!!
Posts: 408
|
Post by tubby on Apr 26, 2015 15:20:46 GMT -5
Tony your measurements seem to confirm the issue. Since mine does not exist the same beahior I wonder if it an issue with the full version. Or perhaps few 3.0 but only if a new filter set is leaded on 3.0 is in use. I will try and get full and get time to do some testing. I have been looking for an excuse anyway.
I think the +10db on the LFE is normal as I don't think REW plays the tone at -10 compared to the main channels so the xmc adds +10 as it is supposed to to the LFE signal. REW does not know it is an LFE channel. It just plays the same signal regardless of the channel.
|
|
|
Post by petes on Apr 26, 2015 16:09:06 GMT -5
Nick -- I did some extensive testing this morning and here is what I found. All tests are measured from the XMC-1 XLR outputs. (no speakers, no microphone)
1. The crossover ignores the gain structure set by dirac for the sub and appears to apply the gain structure of the mains to the sub instead. 2. Filters from the mains and the sub are applied to audio sent to the sub from the crossover. 2a. You can stop this by setting the curtain of the speaker that is being crossed over to the crossover point. (this will only allow the subs filter to affect the response)
NOTE -- These graphs do not measure the same way that speaker measurements would. There is no room gain and no speakers involved. The higher gain of the LFE measurement would be normal since it requires more gain to get the subwoofer to play the same SPL as the mains. (My Subs are about 6db less efficient then the mains) Also, the crossovers would not match correctly due to this. However, I did find that the LFE is about 10db hot compared to when they play from the LFE then played from the xover. Not sure why this is like this and I don't have another branded receiver to test with. Below is the graphs of the original filter. (no custom curtains applied)
Blue = INPUT - LFE OUTPUT - SUBWOOFER Red = INPUT - FRONT OUTPUT - SUBWOOFER Purple = INPUT - FRONT OUTPUT - SUBWOOFER Yellow = INPUT - FRONT OUTPUT - FRONT View Attachment Below is the graphs of new filter with curtain set at the crossover point. (curtains were not changed for the subwoofer.)
Purple = INPUT - LFE OUTPUT - SUBWOOFER Green = INPUT - FRONT OUTPUT - SUBWOOFER Blue = INPUT - FRONT OUTPUT - FRONT View Attachment
REW file
Tony Hmm - that sounds worrying - I guess the fact that the crossovers and the filters operate "indepently" - i.e. the crossovers know nothing about the filters and the filters know nother about the crossovers - make this quite likely (but wrong). If the crossover literally has nothing to do with the filters, this would seem to suggest that the filters are applied first, and then the crossovers are then applied to this gained result. What's the downside of setting the curtains close to the crossover point - wouldn't you want to do this anyway otherwise you're going to be creating filters to try and correct ranges the mains can't really reproduce. Or, should we maybe be looking at where Dirac thinks the response of the speaker begins to degrade and set the crossover to match that - i.e. my speakers crossover according to the docs at 90Hz, but the Dirac response post filter is flat to my target to around 70Hz. So, maybe set both the curtains and the crossover to 70Hz. I guess the broader question is, since the filter and crossovers don't know about each other, how do we best integrate the two?
|
|
|
Post by unsound on Apr 26, 2015 17:41:53 GMT -5
Speaking of filters and crossovers, I just noticed on Dirac LE, that it applies a low pass on the subs based on the crossover frequency selected for the mains. Does that mean LFE is routed to mains above crossover? Or, it still goes to the sub but no correction is applied?
|
|
|
Post by unsound on Apr 26, 2015 17:59:32 GMT -5
Speaking of filters and crossovers, I just noticed on Dirac LE, that it applies a low pass on the subs based on the crossover frequency selected for the mains. Does that mean LFE is routed to mains above crossover? Or, it still goes to the sub but no correction is applied? Re-ran Dirac with socketman's CSL file and it's definitely not the case now.
|
|
|
Post by markc on Apr 27, 2015 1:09:04 GMT -5
It does very much look like the crossover to the sub works after Dirac processing. This means that the whole frequency range of the channels are altered by Dirac, then, for the "small" speakers, the crossover extracts bass and sends it to the sub woofer where it is summed to the LFE. Tony, your graph here is pretty conclusive as the red line contains the Dirac adjustments from both the purple and blue lines and you may have confirmed a new XMC-1 blooper.Well done to Nickwin for noticing that there was an audible problem.What MUST happen for any room correction system (which is actually an "in room speaker response correction" rather than a room correction hence the Dirac needs to occur last in the chain) with "small" speakers, is that the untouched full range channels should have bass extracted from the small speaker channels, these are summed to LFE, THEN the Dirac subwoofer filter "corrects" this summed LFE+bass to compensate for the in room subwoofer response whilst the main channel minus the crossed over bass has the Dirac filter for that channel applied.It really doesn't look like this is happening correctly on the XMC and this could also explain why the initial reports of Dirac after it was released was given complaints of the bass being sucked out and the trebles being boosted - it may not be just down to the frequency response and calibration file of the microphone. I haven't got time now to look over those old posts to see if it applies even back then.If the behaviour of the XMC is to apply Dirac and then extract bass to the LFE, then this is nothing to do with stale filters or stale measurements but an inherent error that will persist whatever filter set is used. Let us hope this is not the case, or at least, let us hope that this can be corrected in firmware meaning that when playing audio, the crossover and extracted bass/LFE summation occurs before Dirac filter implementation.
|
|
|
Post by petes on Apr 27, 2015 2:31:48 GMT -5
Hmm - that sounds worrying - I guess the fact that the crossovers and the filters operate "indepently" - i.e. the crossovers know nothing about the filters and the filters know nother about the crossovers - make this quite likely (but wrong). If the crossover literally has nothing to do with the filters, this would seem to suggest that the filters are applied first, and then the crossovers are then applied to this gained result. What's the downside of setting the curtains close to the crossover point - wouldn't you want to do this anyway otherwise you're going to be creating filters to try and correct ranges the mains can't really reproduce. Or, should we maybe be looking at where Dirac thinks the response of the speaker begins to degrade and set the crossover to match that - i.e. my speakers crossover according to the docs at 90Hz, but the Dirac response post filter is flat to my target to around 70Hz. So, maybe set both the curtains and the crossover to 70Hz. I guess the broader question is, since the filter and crossovers don't know about each other, how do we best integrate the two? So, thinking aboiut this overnight .... Setting the curtains so that the main filters won't getting applied to the signal which is going to be cross'ed over to the sub only solves half the problem - it won't have the main gains/cuts applied, but of course it also won't have the gains/cuts it should have applied from the sub filters. So - I think this is probably one we should ask the simple questoin to Keith and Co. - KeithL , does the routing with Dirac go Source-Dirac-Crossover-Output, or Source-Crossover-Dirac-Output. Maybe we can get clarity or put this one to bed before it mushrooms - unless it needs to of course. TBH, I'm delighted with the sound from my XMC, but of course now I'm back to tinkering. pete S
|
|
|
Post by nickwin on Apr 27, 2015 9:25:27 GMT -5
Thanks for picking up the baton Tony. Its good to know that its not just my unit (or me) going crazy. I had wondered numerous times if the way I was setting up my target curves could could be responsible for my changing sub response and this explains it. I was moving the curtain and as we can see now from Tonys measurement, this will effect the routed bass response.
Petes, in general the ideal (XMC1 oddities aside) when EQing two speakers before applying a crossover is to eq the mains a full octave below and the sub a full octave above the crossover (assuming they have good output in that range). Each speaker will effect the summed response above and below the crossover point, so starting with both the them EQed flat ensures the the crossover results in a smooth summed response. Remember that with the 12db an octave filter the XMC applies to the mains, they will only be down 12db at 40z with a typical 80z crossover. This might not be the best practice with the XMC though as it has a whole new set of variables to consider. At the moment it looks like moving the curtain to the crossover point might be the way to go.
I haven't put much thought into this yet, but it doesn't seem like it has to be this way even considering that Dirac has nothing to do with the crossovers. Couldn't it go: input>Crossover>Dirac (filters and gain)> delay? So that the gain, filters and delay applied to each speaker was actually intended for that speaker? If this allowed for a user adjustable delay setting that would be a win win.
I'm not postive but I think the 10db boost on the LFE channel is standard to consumer pro/pro's.
Ray said in another thread that Delay was the last thing applied in the DSP, so it seems kind of odd that other setting set by Dirac, the filters and gain, would be applied before. Based on what I saw in my speaker measurements it looked like the Dirac HPF's were also applied after the crossover because my routed bass didn't seem to be significantly reduced below the HPF setting when using one.
|
|
|
Post by nickwin on Apr 27, 2015 10:43:57 GMT -5
I still wonder if things are functioning 100% consistently. As far as I know all of the graphs I posted on this thread were with the same, or very very close, HPF and curtain settings, yet I got all these varying results. Right now my mains curtains are set to 45z with a 90z crossover, yet as you can see in the last graph I posted the routed bass appears to look as it should at the moment. Then you also have tubby's measurements that look like they should as well. It seems crazy but its like the way it handles routed bass isn't consistent. I guess it is a computer so anything is possible...
|
|
|
Post by nickwin on Apr 27, 2015 10:59:51 GMT -5
Hmm - that sounds worrying - I guess the fact that the crossovers and the filters operate "indepently" - i.e. the crossovers know nothing about the filters and the filters know nother about the crossovers - make this quite likely (but wrong). If the crossover literally has nothing to do with the filters, this would seem to suggest that the filters are applied first, and then the crossovers are then applied to this gained result. What's the downside of setting the curtains close to the crossover point - wouldn't you want to do this anyway otherwise you're going to be creating filters to try and correct ranges the mains can't really reproduce. Or, should we maybe be looking at where Dirac thinks the response of the speaker begins to degrade and set the crossover to match that - i.e. my speakers crossover according to the docs at 90Hz, but the Dirac response post filter is flat to my target to around 70Hz. So, maybe set both the curtains and the crossover to 70Hz. I guess the broader question is, since the filter and crossovers don't know about each other, how do we best integrate the two? So, thinking aboiut this overnight .... Setting the curtains so that the main filters won't getting applied to the signal which is going to be cross'ed over to the sub only solves half the problem - it won't have the main gains/cuts applied, but of course it also won't have the gains/cuts it should have applied from the sub filters. So - I think this is probably one we should ask the simple questoin to Keith and Co. - KeithL , does the routing with Dirac go Source-Dirac-Crossover-Output, or Source-Crossover-Dirac-Output. Maybe we can get clarity or put this one to bed before it mushrooms - unless it needs to of course. TBH, I'm delighted with the sound from my XMC, but of course now I'm back to tinkering. pete S The way I am understanding it everything within the curtain of the main is applied to the sub on top of the subs filters. So by setting the curtain of the main right at the crossover point, the sub will get only the filters for the sub. I'm hoping Lonnie or Ray will chime in here, I bet they know whats going on. They should be back from Axpona this week. I agree that it sounds good as is BTW. From what were seeing it should be capable of being even better though. Specifically the routed bass.
|
|
|
Post by bestboy on Apr 27, 2015 12:46:52 GMT -5
I'm not postive but I think the 10db boost on the LFE channel is standard to consumer pro/pro's. The LFE channel is recorded with a level offset of –10 dB. This offset is compensated for in the reproduction system, where the LFE loudspeaker has an acoustic output (within its low frequency passband) of +10 dB with respect to the other channels.
|
|
|
Post by markc on Apr 27, 2015 13:28:04 GMT -5
My issue is the looming April 30th Cut off for the $500 price increase for the XMC-1 plus the lack of upgrade for life on purchasing after that date.
I have been sitting on the fence regarding the XMC-1 as I live in Europe and don't want to pay import duties and $200 shipping on something that could have hardware problems too. In addition, I have been waiting for any bugs to be identified and fixed but so far there are quite a few outstanding ones, plus here a new major discovery that, if true, is not something I want from a $2500 DAC/pre-amp/HDMI switcher, which are currently the working functions of the XMC-1.
The XMC is barely meeting the title of a processor for my needs as the processing part seems broken - firstly with over-extravagant tinkering with the microphone measurements (and repeated re-measurements)and calibration files to get something that passes sensible scrutiny in Dirac setup, and now with an almost certain error with the way the summed bass from the "small" speakers which is destined to be routed to the subwoofer is first equalised with the Dirac filters for the satellite speaker on the channel that the redirected bass originated.
I have three days to decide whether to buy a £2000 DAC/pre-amp/HDMI switcher in that case.
I know that there will be calls from satisfied owners saying that even if Dirac is not doing what it should, it is still a good purchase and sounds great.
I am conscious that there is a big case of the Emperors New Clothes and everyone wanting it to sound great with each revelation about what we/the XMC was doing wrong before.
On the XMC-1's debut, everyone felt the XMC was great, then Dirac came out and the forum reports was that it sounded even better.
Then we found out that what people were hearing as sounding better with Dirac was probably just sounding different and was technically wrong and the microphone and calibration file and one size fits all target curve are still not doing something that a purist system should do but were instead emphasising high frequencies and dulling the low frequencies.
Then, by applying Tony Ansat's calibration and house curves etc it could be made to sound properly better but now we find that it isn't better at all, but just sounds different, yet again, and incorrect at that.
It seems the XMC is making some basic errors in the way it is handling multichannel audio with any more than a 2.0 system.
Again, I'll moot a wonder if this summed redirected bass problem is what has been the issue with Dirac all along, and the microphone calibration was a misdirection.
Tony, I want to clarify what I am interpreting from your first chart that I reposted above, in large.
The gold/yellow line is the front channel output with speaker set to small and the crossover active(set to around 80Hz I guess)and Dirac filters applied causing the waves in the trace. The purple is the front channel bass extracted using the crossover to be redirected to the subwoofer and showing the gold front channels characteristic Dirac filter cuts/boosts (but without the crossover filter downslope like in the gold trace.) The teal blue line is the LFE channel(not subwoofer - just the .1) showing the measured Dirac filter cuts/boosts for the subwoofer speaker unit. The red line is the subwoofer output containing summed redirected bass(ie the puple) plus LFE (teal) with the Dirac filter characteristics of BOTH the teal blue and the gold lines.
The red trace is what you hear in a system that has any of the main channels set to "small" with a crossover active and a subwoofer present. Extracted bass plus LFE plus incorrect equalisation.
Unaffected people would then be in two groups: Those with no subwoofer and then those with a subwoofer, but all main speakers set to "large" so there is no redirected bass. Are these the people for whom Dirac is working and reporting as happy?
Finally, the subwoofer output jacks from the back of the XMC should be 10dB COLDER than the main channels if the speakers are set to large and 15dB COLDER than the main channels if ANY of the speakers are set to small.
This is the specification of LFE which has a wider dynamic range and would clip the standard RCA output voltage. Gain must be applied external to the receiver (subwoofer volume control) with 10dB gain or 15dB gain. The 15db is because LFE + summed bass into the subwoofer out leads would be more likely to clip so it is further attenuated 5db internally and requires external boost.
|
|
|
Post by nickwin on Apr 27, 2015 14:13:46 GMT -5
My issue is the looming April 30th Cut off for the $500 price increase for the XMC-1 plus the lack of upgrade for life on purchasing after that date. I have been sitting on the fence regarding the XMC-1 as I live in Europe and don't want to pay import duties and $200 shipping on something that could have hardware problems too. In addition, I have been waiting for any bugs to be identified and fixed but so far there are quite a few outstanding ones, plus here a new major discovery that, if true, is not something I want from a $2500 DAC/pre-amp/HDMI switcher, which are currently the working functions of the XMC-1. The XMC is barely meeting the title of a processor for my needs as the processing part seems broken - firstly with over-extravagant tinkering with the microphone measurements (and repeated re-measurements)and calibration files to get something that passes sensible scrutiny in Dirac setup, and now with an almost certain error with the way the summed bass from the "small" speakers which is destined to be routed to the subwoofer is first equalized with the Dirac filters for the satellite speaker on the channel that the redirected bass originated. I have three days to decide whether to buy a £2000 DAC/pre-amp/HDMI switcher in that case. I know that there will be calls from satisfied owners saying that even if Dirac is not doing what it should, it is still a good purchase and sounds great. I am conscious that there is a big case of the Emperors New Clothes and everyone wanting it to sound great with each revelation about what we/the XMC was doing wrong before. On the XMC-1's debut, everyone felt the XMC was great, then Dirac came out and the forum reports was that it sounded even better. Then we found out that what people were hearing as sounding better with Dirac was probably just sounding different and was technically wrong and the microphone and calibration file and one size fits all target curve are still not doing something that a purist system should do but were instead emphasising high frequencies and dulling the low frequencies. Then, by applying Tony Ansat's calibration and house curves etc it could be made to sound properly better but now we find that it isn't better at all, but just sounds different, yet again, and incorrect at that. It seems the XMC is making some basic errors in the way it is handling multichannel audio with any more than a 2.0 system. Again, I'll moot a wonder if this summed redirected bass problem is what has been the issue with Dirac all along, and the microphone calibration was a misdirection. Tony, I want to clarify what I am interpreting from your first chart that I reposted above, in large. The gold/yellow line is the front channel output with speaker set to small and the crossover active(set to around 80Hz I guess)and Dirac filters applied causing the waves in the trace. The purple is the front channel bass extracted using the crossover to be redirected to the subwoofer and showing the gold front channels characteristic Dirac filter cuts/boosts (but without the crossover filter downslope like in the gold trace.) The teal blue line is the LFE channel(not subwoofer - just the .1) showing the measured Dirac filter cuts/boosts for the subwoofer speaker unit. The red line is the subwoofer output containing summed redirected bass(ie the puple) plus LFE (teal) with the Dirac filter characteristics of BOTH the teal blue and the gold lines. The red trace is what you hear in a system that has any of the main channels set to "small" with a crossover active and a subwoofer present. Extracted bass plus LFE plus incorrect equalisation. Unaffected people would then be in two groups: Those with no subwoofer and then those with a subwoofer, but all main speakers set to "large" so there is no redirected bass. Are these the people for whom Dirac is working and reporting as happy? Finally, the subwoofer output jacks from the back of the XMC should be 10dB COLDER than the main channels if the speakers are set to large and 15dB COLDER than the main channels if ANY of the speakers are set to small. This is the specification of LFE which has a wider dynamic range and would clip the standard RCA output voltage. Gain must be applied external to the receiver (subwoofer volume control) with 10dB gain or 15dB gain. The 15db is because LFE + summed bass into the subwoofer out leads would be more likely to clip so it is further attenuated 5db internally and requires external boost. You make some really good points and I agree for the most part but I just want to say a couple things. The cal files are quite different on there own and will result in a noticeably different sound depending on which one you use. I had my mic calibrated at 0, 45 and 90 degrees and as far as I can tell the cal file provided by Emotive isn't an accurate any degree file, it doesn't look like any of mine. Emotiva's stance is an accurate 90 deg file isn't "correct" in this application but they are the minority with that opinion, Dirac themselves recommend using a 90 deg orientation. If you want a 90 cal you can pay to have your mic calibrated or use ansats file, which will be pretty close to accurate. Anyways, its definitely a separate issue, but its an easy one to work around. Same thing on the lack of user adjustable delay (to fine tune phase at the crossover), a big oversight IMO, but easy to get around with a MiniDSP. In all fairness I should say that as is my system sounds fantastic. Dirac aside I don't know of any processor that can match the XMC for sound quality at its price. At least not for the $2000 $2500 it sells for in the US. In general Dirac has been super consistent and easy to use which is something I can't say about Audyssey. Being able to fine tune the target curve is pretty amazing as well. If what were seeing here is real then it needs to be corrected and the XMC will be even better, and what it should have been from the beginning. The UMC1 had a very similar issue, actually maybe the exact same issue, early on and it was corrected in FW. Hopefully Emotiva can do the same thing here.
|
|
|
Post by nickwin on Apr 27, 2015 14:26:16 GMT -5
If anyone is doing any more testing could they check present 1 or 2? Im wondering if the PEQ has the same behavior.
Tony, did you by any chance test those during your measurements yesterday?
|
|
|
Post by navybuck on Apr 27, 2015 14:38:48 GMT -5
My issue is the looming April 30th Cut off for the $500 price increase for the XMC-1 plus the lack of upgrade for life on purchasing after that date. I have been sitting on the fence regarding the XMC-1 as I live in Europe and don't want to pay import duties and $200 shipping on something that could have hardware problems too. In addition, I have been waiting for any bugs to be identified and fixed but so far there are quite a few outstanding ones, plus here a new major discovery that, if true, is not something I want from a $2500 DAC/pre-amp/HDMI switcher, which are currently the working functions of the XMC-1. The XMC is barely meeting the title of a processor for my needs as the processing part seems broken - firstly with over-extravagant tinkering with the microphone measurements (and repeated re-measurements)and calibration files to get something that passes sensible scrutiny in Dirac setup, and now with an almost certain error with the way the summed bass from the "small" speakers which is destined to be routed to the subwoofer is first equalised with the Dirac filters for the satellite speaker on the channel that the redirected bass originated. I have three days to decide whether to buy a £2000 DAC/pre-amp/HDMI switcher in that case. I know that there will be calls from satisfied owners saying that even if Dirac is not doing what it should, it is still a good purchase and sounds great. I am conscious that there is a big case of the Emperors New Clothes and everyone wanting it to sound great with each revelation about what we/the XMC was doing wrong before. On the XMC-1's debut, everyone felt the XMC was great, then Dirac came out and the forum reports was that it sounded even better. Then we found out that what people were hearing as sounding better with Dirac was probably just sounding different and was technically wrong and the microphone and calibration file and one size fits all target curve are still not doing something that a purist system should do but were instead emphasising high frequencies and dulling the low frequencies. Then, by applying Tony Ansat's calibration and house curves etc it could be made to sound properly better but now we find that it isn't better at all, but just sounds different, yet again, and incorrect at that. It seems the XMC is making some basic errors in the way it is handling multichannel audio with any more than a 2.0 system. Again, I'll moot a wonder if this summed redirected bass problem is what has been the issue with Dirac all along, and the microphone calibration was a misdirection. Tony, I want to clarify what I am interpreting from your first chart that I reposted above, in large. The gold/yellow line is the front channel output with speaker set to small and the crossover active(set to around 80Hz I guess)and Dirac filters applied causing the waves in the trace. The purple is the front channel bass extracted using the crossover to be redirected to the subwoofer and showing the gold front channels characteristic Dirac filter cuts/boosts (but without the crossover filter downslope like in the gold trace.) The teal blue line is the LFE channel(not subwoofer - just the .1) showing the measured Dirac filter cuts/boosts for the subwoofer speaker unit. The red line is the subwoofer output containing summed redirected bass(ie the puple) plus LFE (teal) with the Dirac filter characteristics of BOTH the teal blue and the gold lines. The red trace is what you hear in a system that has any of the main channels set to "small" with a crossover active and a subwoofer present. Extracted bass plus LFE plus incorrect equalisation. Unaffected people would then be in two groups: Those with no subwoofer and then those with a subwoofer, but all main speakers set to "large" so there is no redirected bass. Are these the people for whom Dirac is working and reporting as happy? Finally, the subwoofer output jacks from the back of the XMC should be 10dB COLDER than the main channels if the speakers are set to large and 15dB COLDER than the main channels if ANY of the speakers are set to small. This is the specification of LFE which has a wider dynamic range and would clip the standard RCA output voltage. Gain must be applied external to the receiver (subwoofer volume control) with 10dB gain or 15dB gain. The 15db is because LFE + summed bass into the subwoofer out leads would be more likely to clip so it is further attenuated 5db internally and requires external boost. From reading your post, It doesn't appear that you are on the fence at all, so your decision should be easy.
|
|
|
Post by socketman on Apr 27, 2015 16:36:19 GMT -5
NavyBuck
I am in the group of people who have 2 subs and everything sounds great. Even though it seems the redirected bass is being corrected as though it is being played by the main speakers I don't find this objectionable . I would think redirected bass would have the sub/LFE filters applied after crossover. As always shipping something internationally from ID company will always carry risk but I think the risk is low. I am not as far away as you are but far enough that shipping it back is expensive and possibly cost prohibitive. There are some owners in Australia , Micronesia and the Netherlands and honestly the failure rate is pretty low overall from what has been mentioned on the forum. Hopefully they will allow separate parts purchases .I now there is at least one other member on the board that does repairs overseas so I would not worry too much about that.
Richard
|
|
|
Post by nickwin on Apr 27, 2015 19:18:26 GMT -5
I'm pretty sure the 10db boost on the LFE is normal. This is from Hometheaterhifi.com: "Stop! Before you run home and set your subwoofer 10 dB higher than youÕve already set it, you need to read on and find out why this is not necessary. For consumers, we want things to be simple. Up until recently, only serious enthusiasts would use an SPL meter to set the levels of their home theater equipment, let alone ask everyone to remember to set the LFE channel differently. For this reason, home Dolby Digital equipment is pre-set to play LFE data 10 dB higher than a main channel (or 10dB higher than the bass from a main channel). It is only necessary to set the subwoofer relative to a main channel and the LFE level will be correct. Very few processors allow direct manipulation of the LFE level. That is why the menus on most receivers say "Subwoofer Level" rather than "LFE Level". Here is a link to the full article which explains it in detail: www.hometheaterhifi.com/volume_7_2/feature-article-misunderstood-lfe-channel-april-2000.html
|
|
|
Post by Dan Laufman on Apr 27, 2015 19:42:16 GMT -5
nickwin Don't purchase the XMC-1 if you feel pressured. We wan't you to be happy with us and I don't think you are.
Maybe another brand would be a better fit.
We can't be all things to all people.
Maybe next time. Cheers. Dan
|
|
|
Post by nickwin on Apr 27, 2015 19:55:13 GMT -5
nickwin Don't purchase the XMC-1 if you feel pressured. We wan't you to be happy with us and I don't think you are. Maybe another brand would be a better fit. We can't be all things to all people. Maybe next time. Cheers. Dan Hey Dan, are you sure that was suppose to be directed at me? I already own the XMC1 and am happy with it. I think its a fantastic product overall. Ive said that numerous times in this very thread. I actually just PM'ed someone earlier today about why the they should buy an XMC1. That said what I and others have documented in this thread seems to be somewhat of a flaw. I only bring it up because Im looking for a work around, honestly I wasn't expecting such a negative response. Ive been waiting to hear back from Emotiva support since last Friday, I understand you guys were busy at Axpona... that said Im not sure how Im suppose to feel about the response to my legitimate inquiry being essentially "go elsewhere", from the CEO no less The XMC1 is great but I want it to be as good as it can be. I feel like thats good for everybody. None of this was intended to be an attack, the opposite actually Im trying to help myself and others get the most out of the XMC. I was an early adopter of the UMC1 and recall it having a similar problem early on and after people pointed it out here you guys fixed it with a FW update. I thought that was really cool. It made the UMC1 a better product, I still use mine to this day. Congrats on the generally very positive Hometheaterhifi.com review, but in my defense Robert Kozel said the same thing about the routed bass as I did, "With the Dirac Speaker Preset enabled, I noticed that the bass response was not to my liking, and I felt that the vocals were a bit hard to hear. I ran another set of Dirac measurements, and this time I removed the subwoofers from the system and just told the XMC-1 that I had full-range speakers since my towers have built-in subs. The result was dramatically improved bass response, at least in my configuration." Nick
|
|
|
Post by ansat on Apr 27, 2015 21:17:49 GMT -5
My issue is the looming April 30th Cut off for the $500 price increase for the XMC-1 plus the lack of upgrade for life on purchasing after that date. I have been sitting on the fence regarding the XMC-1 as I live in Europe and don't want to pay import duties and $200 shipping on something that could have hardware problems too. In addition, I have been waiting for any bugs to be identified and fixed but so far there are quite a few outstanding ones, plus here a new major discovery that, if true, is not something I want from a $2500 DAC/pre-amp/HDMI switcher, which are currently the working functions of the XMC-1. The XMC is barely meeting the title of a processor for my needs as the processing part seems broken - firstly with over-extravagant tinkering with the microphone measurements (and repeated re-measurements)and calibration files to get something that passes sensible scrutiny in Dirac setup, and now with an almost certain error with the way the summed bass from the "small" speakers which is destined to be routed to the subwoofer is first equalised with the Dirac filters for the satellite speaker on the channel that the redirected bass originated. I have three days to decide whether to buy a £2000 DAC/pre-amp/HDMI switcher in that case. I know that there will be calls from satisfied owners saying that even if Dirac is not doing what it should, it is still a good purchase and sounds great. I am conscious that there is a big case of the Emperors New Clothes and everyone wanting it to sound great with each revelation about what we/the XMC was doing wrong before. On the XMC-1's debut, everyone felt the XMC was great, then Dirac came out and the forum reports was that it sounded even better. Then we found out that what people were hearing as sounding better with Dirac was probably just sounding different and was technically wrong and the microphone and calibration file and one size fits all target curve are still not doing something that a purist system should do but were instead emphasising high frequencies and dulling the low frequencies. Then, by applying Tony Ansat's calibration and house curves etc it could be made to sound properly better but now we find that it isn't better at all, but just sounds different, yet again, and incorrect at that. It seems the XMC is making some basic errors in the way it is handling multichannel audio with any more than a 2.0 system. Again, I'll moot a wonder if this summed redirected bass problem is what has been the issue with Dirac all along, and the microphone calibration was a misdirection. Tony, I want to clarify what I am interpreting from your first chart that I reposted above, in large. The gold/yellow line is the front channel output with speaker set to small and the crossover active(set to around 80Hz I guess)and Dirac filters applied causing the waves in the trace. The purple is the front channel bass extracted using the crossover to be redirected to the subwoofer and showing the gold front channels characteristic Dirac filter cuts/boosts (but without the crossover filter downslope like in the gold trace.) The teal blue line is the LFE channel(not subwoofer - just the .1) showing the measured Dirac filter cuts/boosts for the subwoofer speaker unit. The red line is the subwoofer output containing summed redirected bass(ie the puple) plus LFE (teal) with the Dirac filter characteristics of BOTH the teal blue and the gold lines. The red trace is what you hear in a system that has any of the main channels set to "small" with a crossover active and a subwoofer present. Extracted bass plus LFE plus incorrect equalisation. Unaffected people would then be in two groups: Those with no subwoofer and then those with a subwoofer, but all main speakers set to "large" so there is no redirected bass. Are these the people for whom Dirac is working and reporting as happy? Finally, the subwoofer output jacks from the back of the XMC should be 10dB COLDER than the main channels if the speakers are set to large and 15dB COLDER than the main channels if ANY of the speakers are set to small. This is the specification of LFE which has a wider dynamic range and would clip the standard RCA output voltage. Gain must be applied external to the receiver (subwoofer volume control) with 10dB gain or 15dB gain. The 15db is because LFE + summed bass into the subwoofer out leads would be more likely to clip so it is further attenuated 5db internally and requires external boost. I made a correction to my original post on this -- I had the purple line mis-labeled. I will start with the calibration -- The calibration file issue can be summed up really easy - Emotiva designed the calibration file to sound good in their test systems. My calibration file is a "What you see is what you get". correct. this was mis-labeled - This is a full range measurement of my main. correct. The red shows the characteristics of both the subwoofer output and the main full range output. The question of right and wrong is another story. And a spot where I struggle. I believe that Dirac is doing exactly what it is supposed to be doing. Dirac's job is to create a single filter per speaker, set delays (based in initial impulse) For the crossover -- From my questions and posts from Ray and Lonnie, I do not believe that emotiva has the ability to access the filters. So this is an all or nothing per speaker. This would mean that this is the only way that a crossover could be applied is by having the filters from each be applied. I have never measured anything to the extent that I have measured the XMC and do not know if Audyssey exhibits the same phenomenon. But I suspect that any technology that creates a filter that is the inverse of the response would act like this. But one key point in my original post to remember is that by setting the curtain to the crossover, this can be bypassed. This would place no corrections below the crossover and would not get summed into what is sent to the sub. This would be the equivalent of Dirac knowing what the crossover is. The only aspect where I think that the Dirac / XMC has a problem that cannot be worked around through setup is the phase alignment of multiple subs and phase aligning to the crossover. This is all in the delay adjustment of the subwoofer. This could be added into Dirac or the XMC. But for now, a second processor is the only option for making adjustments. (miniDSP is my current favorite for this.) Tony
|
|