|
Post by monkumonku on May 18, 2010 17:11:20 GMT -5
See Lonnie's post in the "Marcus, EMO, and Locked Threads" thread... I just read it, and it confirms that the EQ and crossover are indeed backwards. Instead of summing the crossover signals and putting it through the sub EQ which would make room correction straightforward, now you have 7 coarse equalizers affecting bass material that bypasses the sub EQ making good room EQ all but impossible. Hmmm.... Huh? Lonnie's explanation said that in two channel stereo the signal below the high pass is summed and sent to the sub. For multi channel, the LFE channel is the one that gets equalized. How does "7 coarse equalizers" enter into the picture?
|
|
LCSeminole
Global Moderator
Res firma mitescere nescit.
Posts: 20,501
|
Post by LCSeminole on May 18, 2010 17:16:58 GMT -5
For those interested or wanting to point others to Lonnie's Bass Management breakdown, I've "Stickied" his words in a posting at the top of the "Processors/Preamps" section. If you have comments about it, please continue to post in this thread. I would like to thank Lonnie for taking time to paraphrase his work on the UMC-1 bass management.
|
|
|
Post by johnnyg on May 18, 2010 17:25:15 GMT -5
I just read it, and it confirms that the EQ and crossover are indeed backwards. Instead of summing the crossover signals and putting it through the sub EQ which would make room correction straightforward, now you have 7 coarse equalizers affecting bass material that bypasses the sub EQ making good room EQ all but impossible. Hmmm.... Huh? Lonnie's explanation said that in two channel stereo the signal below the high pass is summed and sent to the sub. For multi channel, the LFE channel is the one that gets equalized. How does "7 coarse equalizers" enter into the picture? Unless we are reading different posts from Lonnie, I don't see where he mentions anything about two channel stereo vs multi channel modes. What he says is that the sub EQ only applies to the incoming LFE signal (which only exists with multichannel source material), but when you have the crossovers enabled you could have as many as 8 sources feeding the subwoofer (7.1) - each with its own EQ affecting the summed response. "3. When you set a high pass crossover for each of your speakers, everything below the crossover frequency will be summed together and sent to the subs. So the EQ settings (for each speaker) below the crossover frequency will apply to the summed bass response." - Lonnie
|
|
|
Post by nickwin on May 18, 2010 17:27:55 GMT -5
I just read it, and it confirms that the EQ and crossover are indeed backwards. Instead of summing the crossover signals and putting it through the sub EQ which would make room correction straightforward, now you have 7 coarse equalizers affecting bass material that bypasses the sub EQ making good room EQ all but impossible. Hmmm.... Huh? Lonnie's explanation said that in two channel stereo the signal below the high pass is summed and sent to the sub. For multi channel, the LFE channel is the one that gets equalized. How does "7 coarse equalizers" enter into the picture? The downside of this is that the only bass eq available for anything without a LFE channel is the part of the speaker eqs below the crossover point, and the speaker eqs are way to course to be useful in leveling your subwoofer response. Also, like Johnnyg said, on multi channel soundtracks with an LFE channel, there is still a lot of bass information being routed to the your sub from your mains that will not be adjustable via the LFE eq. I am still a little hazy about the summing of the bass. Say I have my L and R speakers 3db down at 31.5hz and my LFE set to -3bd at 35hz. Does that mean that the summed bass output is going to be -6db at around 30-35hz? Or does it trim 3db from the L and R signal and 3db from the LFE for a net trim of 3db? And what if I have my LFE eq at -6db at 35hz, and my mains flat below the crossover point. Wouldn't that ruin the "balance" of a discrete 5.1 soundtrack (because the sum of the bass may have a flat response, but the bass being routed from the mains is going to be 6db hot compared to the bass from the LFE channel at that freq.)
|
|
ratmice
Emo VIPs
I'm not an actor, but I play one on TV.
Posts: 1,853
|
Post by ratmice on May 18, 2010 17:55:26 GMT -5
Unless we are reading different posts from Lonnie, I don't see where he mentions anything about two channel stereo vs multi channel modes. What he says is that the sub EQ only applies to the incoming LFE signal (which only exists with multichannel source material), but when you have the crossovers enabled you could have as many as 8 sources feeding the subwoofer (7.1) - each with its own EQ affecting the summed response. I think you are reading different posts. A little further down the thread is a reply to a question which pertains to this issue. Unless I am not understanding something (which is very likely). Thanks, Lonnie! I think I have it, but just to make sure...what your #3 means is that if I am running 2-channel stereo and have my sub connected, the fronts play down to the x-over level I set for the fronts, the sub will take over from there down, and the sub EQ is controlled by whatever EQ settings I have chosen for the fronts below that x-over. Correct? (Of course, I am ignoring for simplicity the rolloff slopes.) If so, that's great! If not, educate me on what part I missed... Thanks again, Mark Correct, you got it!
|
|
|
Post by spoolio on May 18, 2010 18:13:38 GMT -5
Huh? Lonnie's explanation said that in two channel stereo the signal below the high pass is summed and sent to the sub. For multi channel, the LFE channel is the one that gets equalized. How does "7 coarse equalizers" enter into the picture? The downside of this is that the only bass eq available for anything without a LFE channel is the part of the speaker eqs below the crossover point, and the speaker eqs are way to course to be useful in leveling your subwoofer response. Also, like Johnnyg said, on multi channel soundtracks with an LFE channel, there is still a lot of bass information being routed to the your sub from your mains that will not be adjustable via the LFE eq. I still a little hazy about the summing of the bass. Say I have my L and R speakers 3db down at 31.5hz and my LFE set to -3bd at 35hz. Does that mean that the summed bass output is going to be -6db at around 30-35hz? Or does it trim 3db from the L and R signal and 3db from the LFE for a net trim of 3db? And what if I have my LFE eq at -6db at 35hz, and my mains flat below the crossover point. Wouldn't that ruin the "balance" of a discrete 5.1 soundtrack (because the sum of the bass may have a flat response, but the bass being routed from the mains is going to be 6db hot compared to the bass from the LFE channel at that freq.) My understanding is that the LFE signal will be trimmed by 3dB and the L/R components (plus any other speaker signal components below their respective crossover frequency) will be trimmed per that channel's filter settings below the crossover frequency. All those signals are then D/A'd and summed (that sum includes the LFE signal) and output to the sub. Your second assertion is essentially correct, if the below crossover parts of the 5 other speaker signals are significant enough to override the LFE track (which will be adjusted at the LFE eq settings). How noticeable this will be is probably dependent on the movie content. I agree that the GEQ bands for low-frequency content from the speakers are a little coarse, but my personal opinion is that most people's biggest audible problems are speaker or room related. If you have a nasty room mode at some frequency, your best solution may be treatment rather than EQing the problem away. EQ will only buy you so much--and for music, the price you pay is non-linear phase response, which creates dispersion of time-domain signals by frequency content. That said, if the UMC has the resources available, more defined low frequency EQ bins might be a nice feature for the front channels, at least, considering how many folks are redirecting bass. This, at least, is probably a must implement feature for the expanded XMC-1 set. And while we're speculating, I'd like lower cutoff crossover settings for the fronts. Some movies throw some serious energy down at the 20ish range, and I don't want to kill my fronts by running them full range. Alternatively, I can just keep coupling a cap into my interconnects, Vandersteen style, but it might be nice to not have to do that. Spoolio
|
|
|
Post by johnnyg on May 18, 2010 18:30:42 GMT -5
I think you are reading different posts. A little further down the thread is a reply to a question which pertains to this issue. Unless I am not understanding something (which is very likely). All I see is Lonnie confirming what I said, that two channel works just like multichannel - anything below the crossover goes through the speaker EQ before being sent to the sub. There is no LFE channel in stereo, but it would go through the sub EQ before being summed if it were multichannel.
|
|
|
Post by nickwin on May 18, 2010 18:31:05 GMT -5
The downside of this is that the only bass eq available for anything without a LFE channel is the part of the speaker eqs below the crossover point, and the speaker eqs are way to course to be useful in leveling your subwoofer response. Also, like Johnnyg said, on multi channel soundtracks with an LFE channel, there is still a lot of bass information being routed to the your sub from your mains that will not be adjustable via the LFE eq. I still a little hazy about the summing of the bass. Say I have my L and R speakers 3db down at 31.5hz and my LFE set to -3bd at 35hz. Does that mean that the summed bass output is going to be -6db at around 30-35hz? Or does it trim 3db from the L and R signal and 3db from the LFE for a net trim of 3db? And what if I have my LFE eq at -6db at 35hz, and my mains flat below the crossover point. Wouldn't that ruin the "balance" of a discrete 5.1 soundtrack (because the sum of the bass may have a flat response, but the bass being routed from the mains is going to be 6db hot compared to the bass from the LFE channel at that freq.) My understanding is that the LFE signal will be trimmed by 3dB and the L/R components (plus any other speaker signal components below their respective crossover frequency) will be trimmed per that channel's filter settings below the crossover frequency. All those signals are then D/A'd and summed (that sum includes the LFE signal) and output to the sub. Your second assertion is essentially correct, if the below crossover parts of the 5 other speaker signals are significant enough to override the LFE track (which will be adjusted at the LFE eq settings). How noticeable this will be is probably dependent on the movie content. I agree that the GEQ bands for low-frequency content from the speakers are a little coarse, but my personal opinion is that most people's biggest audible problems are speaker or room related. If you have a nasty room mode at some frequency, your best solution may be treatment rather than EQing the problem away. EQ will only buy you so much--and for music, the price you pay is non-linear phase response, which creates dispersion of time-domain signals by frequency content. That said, if the UMC has the resources available, more defined low frequency EQ bins might be a nice feature for the front channels, at least, considering how many folks are redirecting bass. This, at least, is probably a must implement feature for the expanded XMC-1 set. And while we're speculating, I'd like lower cutoff crossover settings for the fronts. Some movies throw some serious energy down at the 20ish range, and I don't want to kill my fronts by running them full range. Alternatively, I can just keep coupling a cap into my interconnects, Vandersteen style, but it might be nice to not have to do that. Spoolio In my situation, I have a monster room mode that just happens to peak at right about 35-38hz, which corresponds perfectly with the 35hz LFE eq band. So, I set my LFE eq to -9db at 35hz, and that got the peak down to about +3db (its was 10+db originally). Now say hypothetically I'm watching a move in 5.1. When bass is coming from the LFE channel, my in room response will stay pretty flat in that freq. range because I have that 9db trim on LFE, but then say there is a big burst of bass information from the front speakers in roughly same freq. range. That information gets redirected to the sub, but there is no trim at that freq. when its coming from the main speakers so I get a massive blast of bass that's 9db hot compared to what was coming from the LFE channel, which is a significant difference. So not only do I still have a massive peak at 35hz when the signal was routed from my mains, but the bass information coming from my mains is significantly louder than it should be relative to the LFE channel (and relative to how it was originally mastered) It seems like you would need the same eq bands in both LFE and main speaker eq's, to keep everything balanced the way its suppose to be.
|
|
|
Post by gkarracer on May 18, 2010 21:30:03 GMT -5
If I'm interpreting this correctly it works like this:
For direct analog to analog signals the EQ settings are not used in any way whatsoever as they only apply to digitally processed signals. Simply put the analog direct signals come in and everything above the crossover goes to the corresponding speaker and everything below the crossover goes to the sub. EQ is not applied.
For anything digitally processed (regardless of whether the source input is analog or digital), each individual channel gets its own EQ applied. All 8 individual channels have their own separate EQ curve (i.e. 7.1 channels). First the signal gets processed by the EQ curve and then gets routed to the appropriate speaker. Once again signals above the crossover point go to their respective speaker and everything below the crossover point goes to the sub. Note that this crossover routing happens after the EQ is applied.
Effectively what this means is that any EQ settings below the crossover point are equivalent to the sub EQ for that individual channel.
The so called "sub EQ" is merely the EQ for the LFE channel and is applied only to that channel and before the crossover adjustment. It has no bearing on the summed output information. I completely agree that this band should be relabeled "LFE EQ".
Frankly this makes the true sub EQ highly more configurable than a summed sub EQ as it is applied to each channel individually. On the other hand, it has been indicated the lower frequency EQ values are too course (i.e. not enough available adjustments at those frequencies) to make it truly effective.
|
|
|
Post by marius on May 18, 2010 22:06:50 GMT -5
My understanding is that the LFE signal will be trimmed by 3dB and the L/R components (plus any other speaker signal components below their respective crossover frequency) will be trimmed per that channel's filter settings below the crossover frequency. All those signals are then D/A'd and summed (that sum includes the LFE signal) and output to the sub. Your second assertion is essentially correct, if the below crossover parts of the 5 other speaker signals are significant enough to override the LFE track (which will be adjusted at the LFE eq settings). How noticeable this will be is probably dependent on the movie content. I agree that the GEQ bands for low-frequency content from the speakers are a little coarse, but my personal opinion is that most people's biggest audible problems are speaker or room related. If you have a nasty room mode at some frequency, your best solution may be treatment rather than EQing the problem away. EQ will only buy you so much--and for music, the price you pay is non-linear phase response, which creates dispersion of time-domain signals by frequency content. That said, if the UMC has the resources available, more defined low frequency EQ bins might be a nice feature for the front channels, at least, considering how many folks are redirecting bass. This, at least, is probably a must implement feature for the expanded XMC-1 set. And while we're speculating, I'd like lower cutoff crossover settings for the fronts. Some movies throw some serious energy down at the 20ish range, and I don't want to kill my fronts by running them full range. Alternatively, I can just keep coupling a cap into my interconnects, Vandersteen style, but it might be nice to not have to do that. Spoolio In my situation, I have a monster room mode that just happens to peak at right about 35-38hz, which corresponds perfectly with the 35hz LFE eq band. So, I set my LFE eq to -9db at 35hz, and that got the peak down to about +3db (its was 10+db originally). Now say hypothetically I'm watching a move in 5.1. When bass is coming from the LFE channel, my in room response will stay pretty flat in that freq. range because I have that 9db trim on LFE, but then say there is a big burst of bass information from the front speakers in roughly same freq. range. That information gets redirected to the sub, but there is no trim at that freq. when its coming from the main speakers so I get a massive blast of bass that's 9db hot compared to what was coming from the LFE channel, which is a significant difference. So not only do I still have a massive peak at 35hz when the signal was routed from my mains, but the bass information coming from my mains is significantly louder than it should be relative to the LFE channel (and relative to how it was originally mastered) It seems like you would need the same eq bands in both LFE and main speaker eq's, to keep everything balanced the way its suppose to be. That is how I interpreted lonnies post as well. Not too ideal for me as well.
|
|
|
Post by spoolio on May 19, 2010 0:40:32 GMT -5
In my situation, I have a monster room mode that just happens to peak at right about 35-38hz, which corresponds perfectly with the 35hz LFE eq band. So, I set my LFE eq to -9db at 35hz, and that got the peak down to about +3db (its was 10+db originally). Now say hypothetically I'm watching a move in 5.1. When bass is coming from the LFE channel, my in room response will stay pretty flat in that freq. range because I have that 9db trim on LFE, but then say there is a big burst of bass information from the front speakers in roughly same freq. range. That information gets redirected to the sub, but there is no trim at that freq. when its coming from the main speakers so I get a massive blast of bass that's 9db hot compared to what was coming from the LFE channel, which is a significant difference. So not only do I still have a massive peak at 35hz when the signal was routed from my mains, but the bass information coming from my mains is significantly louder than it should be relative to the LFE channel (and relative to how it was originally mastered) It seems like you would need the same eq bands in both LFE and main speaker eq's, to keep everything balanced the way its suppose to be. Yeah, that would be ideal. They might be up against a filter limitation (i.e., only so many bands implementable per channel) so they had to pick them as best they could. I doubt we'll see this changed in the UMC. In your case, your best option may be to try some kind of trapping to tame that mode--although at 35 Hz, that may take some creativity (and commitment) on your part. Room's got a 15-16' axis and concrete/brick-backed walls or something?
|
|
|
Post by merlinwerks on May 19, 2010 8:07:07 GMT -5
For direct analog to analog signals the EQ settings are not used in any way whatsoever as they only apply to digitally processed signals. Simply put the analog direct signals come in and everything above the crossover goes to the corresponding speaker and everything below the crossover goes to the sub. EQ is not applied. Maybe it's just how you're using the term "direct", I would think the above is true for an analog signal/stereo mode. With an analog signal/direct mode there should be no digital processing i.e. xovers at all, unless the xovers are in the analog domain which I don't believe they are.
|
|
|
Post by doc1963 on May 19, 2010 8:21:30 GMT -5
For direct analog to analog signals the EQ settings are not used in any way whatsoever as they only apply to digitally processed signals. Simply put the analog direct signals come in and everything above the crossover goes to the corresponding speaker and everything below the crossover goes to the sub. EQ is not applied. Maybe it's just how you're using the term "direct", I would think the above is true for an analog signal/stereo mode. With an analog signal/direct mode there should be no digital processing i.e. xovers at all, unless the xovers are in the analog domain which I don't believe they are. I believe your correct. All 2 channel analog inputs in "Direct" mode operate full range. Any other mode requires re-digitizing the incoming analog signal, therefore HP/LP filters and EQs are then utilized and active. The 7.1 inputs are strictly a full range analog path with gain.
|
|
|
Post by Nemesis.ie on May 19, 2010 8:36:19 GMT -5
What might be very useful given the way things are operating would be internal test tones that run EQed (I have suggested that test tones should be like that before).
We would need to be able to set test tones to "redirected mode" so a full-rante pink/white noise tone is sent to the sub and the particular channel, that way we can use the (albeit coarse) frequencies of the GEQ to make adjustments per channel that will be correct with our chosen X-over - Emo-Q should of course be able to do that too, but we need to have a "preset x-over" mode for that to work too.
|
|
|
Post by archnmez on May 19, 2010 8:48:05 GMT -5
Hi All, I called emotiva yesterday and spoke to Lonnie on this subject a few hours before he posted. Anyway, Johnnyg and merlinwerks are correct. I was told in direct there are no crossovers and eqs because that is done digitally. My concerns with the bass implementation are the same as those previously expressed by myself and others here. With the LFE having its own EQ and then all the other channels summed with different EQs what is the best way to get a balanced bass from the sub. It would seem having all different channels with potentially different eqs summed together would be a bass disaster. Do I 1 - Set everthing below my crossover to flat for each channel? 2 - Find a eq that I like and set everthing below my crossover for all 7 channels the same? Another concern, if this is what caused the bass in previous FW to be HOT and current FW adjusted the gain down to compensate, how would that affect the blend between LFE and summed bass? Another words, did they lower the total gain including the LFE(whose gain may have been fine) thereby further reducing the LFE presence compared with the summed bass? Will someone with a good sub, 20hz hitting, comment on the bass coherence.
|
|
|
Post by nickwin on May 19, 2010 10:55:54 GMT -5
Yeah, that would be ideal. They might be up against a filter limitation (i.e., only so many bands implementable per channel) so they had to pick them as best they could. I doubt we'll see this changed in the UMC. In your case, your best option may be to try some kind of trapping to tame that mode--although at 35 Hz, that may take some creativity (and commitment) on your part. Room's got a 15-16' axis and concrete/brick-backed walls or something? Pretty close actually, my room is approx. 18'x15' with brick backed walls on two sides. My bass response is actually pretty flat besides the monster peak in the 35hz range. I really lucked out because that peak matches up just about perfectly with the 35hz eq band on the LFE channel. So, what I've done as of right now is put a 9db trim on the LFE channel, which got the peak down to about +3db on that channel. I then put a 6db trim on my front L and R speakers at 31.5hz (which wasn't perfect, but was surprisingly effective.) That drastically reduced the peak I was getting at around 35hz from the L and R channels. I guess I should put that same 6db trim on my other speakers to keep everything balanced, but I haven't messed around with that yet. All of this was done using DVE's BM test tones and a RS SPL meter. DVE is really nice for this because it has 15-150hz sweeps on all individual channels (including LFE) and a sweep with all channels running simultaneously (for the summed output). So for this particular spl spike I had, the EQ actually helped a lot. It didn't completely "fix" my bass response, to the degree that an outboard eq probably would, but it nearly eliminated the biggest spl spike I had in my room (which made a drastic improvement in the overall sound of my system). On the other hand, if you had a big spike that didn't match any of the UMC's eq bands, then they would be no use at all. I am still concerned about the bass "balance", but I've mentioned that a couple times already. Hopefully Lonnie can give us his thoughts on that.
|
|
kwadswor
Minor Hero
Owner of Truth
Posts: 86
|
Post by kwadswor on May 19, 2010 12:34:14 GMT -5
With the LFE having its own EQ and then all the other channels summed with different EQs what is the best way to get a balanced bass from the sub. I say this only partly tounge-in-cheek, but spend $100 and buy a Behringer Feedback Destroyer DSP1124P. It's a fully parametric EQ with a lot more capability than the built-in graphic EQ. It works on the summed signal going to your sub, and you'll still have it in a few years when you upgrade the processor.
|
|
|
Post by loopinfool on May 19, 2010 14:18:08 GMT -5
Here's another thing to note... I think the "hot bass" issue people complained about after using Emo-Q is (mostly) created by this EQ configuration. Emo-Q tests the small speakers and finds they have very little output below, say, 90Hz, so it boosts the EQ curve for that channel. However, that channel's signal will actually be going out the sub and will now be very, very hot. No offense meant to Lonnie or Emotiva, but I find that to be just plain wrong. As currently implemented, Emo-Q simply can't do the right thing using the equalizers, though it tries to. But it's testing the "wrong" speaker. Assuming there's no way to apply the detailed Sub EQ after the channels are summed (maybe it really is an analog sum?), here are two suggestions on how to deal with this issue: - Have Emo-Q always set the EQ flat below the crossover point (either user-set or detected). This doesn't fully solve the issue, but will help the "out of the box" Emo-Q experience.
- Implement a two-stage equalizer for the non-LFE channels, assuming there's available DSP processor power. One would be a coarse EQ like they have now, and the other could even just be a copy of the current LFE EQ, which would now also get applied to any channel being crossed over to the Sub.
Emo-Q would still want to set the coarse EQ flat below the crossover points.
- LoopinFool
|
|
|
Post by nickwin on May 19, 2010 15:23:02 GMT -5
Here's another thing to note... I think the "hot bass" issue people complained about after using Emo-Q is (mostly) created by this EQ configuration. Emo-Q tests the small speakers and finds they have very little output below, say, 90Hz, so it boosts the EQ curve for that channel. However, that channel's signal will actually be going out the sub and will now be very, very hot. No offense meant to Lonnie or Emotiva, but I find that to be just plain wrong. As currently implemented, Emo-Q simply can't do the right thing using the equalizers, though it tries to. But it's testing the "wrong" speaker. Assuming there's no way to apply the detailed Sub EQ after the channels are summed (maybe it really is an analog sum?), here are two suggestions on how to deal with this issue: - Have Emo-Q always set the EQ flat below the crossover point (either user-set or detected). This doesn't fully solve the issue, but will help the "out of the box" Emo-Q experience.
- Implement a two-stage equalizer for the non-LFE channels, assuming there's available DSP processor power. One would be a coarse EQ like they have now, and the other could even just be a copy of the current LFE EQ, which would now also get applied to any channel being crossed over to the Sub.
Emo-Q would still want to set the coarse EQ flat below the crossover points.
- LoopinFool My subwoofer seems to be running hot even when I use the UMC's internal test tones and a SPL meter.
|
|
|
Post by loopinfool on May 19, 2010 16:20:10 GMT -5
My subwoofer seems to be running hot even when I use the UMC's internal test tones and a SPL meter. You're not the only one and that seems to be a different issue. Sorry for re-hashing in this thread. I missed that big Markus thread when catching up after a vacation. - LoopinFool
|
|