|
|
Post by megash0n on May 3, 2020 11:14:02 GMT -5
He's using ASIO drivers in REW which typically allows you to specify the sub channel which should let him do a full sweep. Otherwise, I agree.. Swap some cables and run a sweep that way. It just sucks because you ideally should be adjusting distance and all that. This is going to matter more though when integrating to the mains. It does seem like some setting in REW or the minidsp logically. Using my left and right sub for heights so I'll be running both subs off center and through a minidsp. Hoping to use the mini strictly for time alignment and then see what Dirac will do with frequency response. Cool. If you have not purchased yet, I would look at the HD version at a minimum. Unfortunately, it doesn't offer balanced input, but it gives you more time delay than the non HD version. |
|
|
|
Post by foggy1956 on May 3, 2020 11:18:57 GMT -5
Using my left and right sub for heights so I'll be running both subs off center and through a minidsp. Hoping to use the mini strictly for time alignment and then see what Dirac will do with frequency response. Cool. If you have not purchased yet, I would look at the HD version at a minimum. Unfortunately, it doesn't offer balanced input, but it gives you more time delay than the non HD version. Got the 2x4 advanced which gives me enough time. I may purchase a minidsp with Dirac capability dependent on how things work out. Only use the subs for movies, music is happiest in reference stereo to my ears. |
|
|
|
Post by megash0n on May 3, 2020 11:22:10 GMT -5
Cool. If you have not purchased yet, I would look at the HD version at a minimum. Unfortunately, it doesn't offer balanced input, but it gives you more time delay than the non HD version. Got the 2x4 advanced which gives me enough time. I may purchase a minidsp with Dirac capability dependent on how things work out. Only use the subs for movies, music is happiest in reference stereo to my ears. I have that one as well in balanced form. I've been looking at the ones that support the new DBC as well just in case Emotiva doesn't support this add-on. Still though, I'm not sure if all of that would play nicely together. We'll figure all this out in the coming months. |
|
|
|
Post by kahodges1721 on May 3, 2020 11:30:01 GMT -5
I think I’ve got it worked out (ish) if I set the fronts to large it opens up LFE for center sub. Keep center to small with a crossover of 200hz. You also need enhanced bass selected (not sure why) doing this will carry the measurements out far enough to get a decent comparison. Not as far as I would like as it still drops down earlier than I need. I really wish they would have an adjustable LFE range like most other brands. This would be very useful for people that like to have subs calibrated above 80hz
|
|
|
|
Post by megash0n on May 3, 2020 11:42:40 GMT -5
I think I’ve got it worked out (ish) if I set the fronts to large it opens up LFE for center sub. Keep center to small with a crossover of 200hz. You also need enhanced bass selected (not sure why) doing this will carry the measurements out far enough to get a decent comparison. Not as far as I would like as it still drops down earlier than I need. I really wish they would have an adjustable LFE range like most other brands. This would be very useful for people that like to have subs calibrated above 80hz At the risk of being misunderstood, this is my perception based on what I'm reading. Not at all implying you don't know these things.. Just a second set of eyes or ears here. And, someone with a better understanding I'm sure will fill in the blanks or correct my wrong statements. LFE and mono are not the same thing. LFE doesn't need adjustment. It is an encoded channel meant for producers to use and goes up to 120hz. Crossovers do not come into play here. Trims do. PEQ does. Etc. Mono is going to take everything around the crossover point you are setting, and based on the slope, attenuate that signal. There's probably a better word than attenuate, but you get my point. If you measure your left channel set to large, you will hear nothing out of the sub. You will hear the sub if you use the enhanced mode because you are telling the processor to play out of the left as large, but below the crossover point... Also out of the sub. It really seems as if you are measuring another channel that is set to small with a crossover between 60-80hz. You continue to mention your center channel. Is this what you are pulling the measurement from? I have looked through the menu again this morning, and the only way I can think to produce what you are seeing is to measure a channel set to small. There is no small or large for the center sub. I don't know how one could even produce what you are seeing if truly measuring the center sub "channel" thru REW. Even though it may be a headache, please try to change your wires so that you left channel output goes to your minidsp and set that channel to large. Run a sweep and see if you still have the same result. If you do, then it almost has to be an issue with REW or a rogue filter in the minidsp. |
|
|
|
Post by kahodges1721 on May 3, 2020 11:59:10 GMT -5
I think I’ve got it worked out (ish) if I set the fronts to large it opens up LFE for center sub. Keep center to small with a crossover of 200hz. You also need enhanced bass selected (not sure why) doing this will carry the measurements out far enough to get a decent comparison. Not as far as I would like as it still drops down earlier than I need. I really wish they would have an adjustable LFE range like most other brands. This would be very useful for people that like to have subs calibrated above 80hz At the risk of being misunderstood, this is my perception based on what I'm reading. Not at all implying you don't know these things.. Just a second set of eyes or ears here. And, someone with a better understanding I'm sure will fill in the blanks or correct my wrong statements. LFE and mono are not the same thing. LFE doesn't need adjustment. It is an encoded channel meant for producers to use and goes up to 120hz. Crossovers do not come into play here. Trims do. PEQ does. Etc. Mono is going to take everything around the crossover point you are setting, and based on the slope, attenuate that signal. There's probably a better word than attenuate, but you get my point. If you measure your left channel set to large, you will hear nothing out of the sub. You will hear the sub if you use the enhanced mode because you are telling the processor to play out of the left as large, but below the crossover point... Also out of the sub. It really seems as if you are measuring another channel that is set to small with a crossover between 60-80hz. You continue to mention your center channel. Is this what you are pulling the measurement from? I have looked through the menu again this morning, and the only way I can think to produce what you are seeing is to measure a channel set to small. There is no small or large for the center sub. I don't know how one could even produce what you are seeing if truly measuring the center sub "channel" thru REW. Even though it may be a headache, please try to change your wires so that you left channel output goes to your minidsp and set that channel to large. Run a sweep and see if you still have the same result. If you do, then it almost has to be an issue with REW or a rogue filter in the minidsp. You are correct LFE and Mono are very different which is my point. LFE should be independent of any crossover and holds content up to 120hz as you stated. In most receivers you can carry this channel low pass filter to as high as 250hz. Unless you are applying filters to the sub channel this doesn’t really matter since setting it above 120 provided no difference (it’s basically just applying double filters since the master is encoded with a filter of 120).This changes though when you run REW you can send any signal over LFE. Running LFE on the XMC center sub should show a crossover of around 150hz and no content from other speakers should be routed by the bass management system as was stated above is the low pass used my emotiva. Problem is it doesn’t. It’s directly linked to the center channel speaker regardless of LFE setting. If I set mains to large to get the LFE setting for the center sub I can change the output on the LFE directly proportionate to the crossover used on the center channel which shouldn’t be the case. Now on the other side of things if I select mono on the center sub I would expect all content below any speakers crossover to be diverted to the sub via bass management system. For this if I send an LFE signal to the XMC it should play 0-150hz if I have the crossover set to 200hz on the center and run a sweep on the center speaker from 10-1000 I would expect information sent to the sub from 10-200 and speaker above 200hz (not exactly as that’s not how filters work but for explanation sake I simplified). This should give me a sub reading from 0-200hz but it doesn’t |
|
|
|
Post by geebo on May 3, 2020 12:01:01 GMT -5
Cool. If you have not purchased yet, I would look at the HD version at a minimum. Unfortunately, it doesn't offer balanced input, but it gives you more time delay than the non HD version. Got the 2x4 advanced which gives me enough time. I may purchase a minidsp with Dirac capability dependent on how things work out. Only use the subs for movies, music is happiest in reference stereo to my ears. The 2x4HD can be converted to Dirac 2 for $199 and a firmware upgrade. |
|
|
|
Post by megash0n on May 3, 2020 12:20:47 GMT -5
At the risk of being misunderstood, this is my perception based on what I'm reading. Not at all implying you don't know these things.. Just a second set of eyes or ears here. And, someone with a better understanding I'm sure will fill in the blanks or correct my wrong statements. LFE and mono are not the same thing. LFE doesn't need adjustment. It is an encoded channel meant for producers to use and goes up to 120hz. Crossovers do not come into play here. Trims do. PEQ does. Etc. Mono is going to take everything around the crossover point you are setting, and based on the slope, attenuate that signal. There's probably a better word than attenuate, but you get my point. If you measure your left channel set to large, you will hear nothing out of the sub. You will hear the sub if you use the enhanced mode because you are telling the processor to play out of the left as large, but below the crossover point... Also out of the sub. It really seems as if you are measuring another channel that is set to small with a crossover between 60-80hz. You continue to mention your center channel. Is this what you are pulling the measurement from? I have looked through the menu again this morning, and the only way I can think to produce what you are seeing is to measure a channel set to small. There is no small or large for the center sub. I don't know how one could even produce what you are seeing if truly measuring the center sub "channel" thru REW. Even though it may be a headache, please try to change your wires so that you left channel output goes to your minidsp and set that channel to large. Run a sweep and see if you still have the same result. If you do, then it almost has to be an issue with REW or a rogue filter in the minidsp. You are correct LFE and Mono are very different which is my point. LFE should be independent of any crossover and holds content up to 120hz as you stated. In most receivers you can carry this channel low pass filter to as high as 250hz. Unless you are applying filters to the sub channel this doesn’t really matter since setting it above 120 provided no difference (it’s basically just applying double filters since the master is encoded with a filter of 120).This changes though when you run REW you can send any signal over LFE. Running LFE on the XMC center sub should show a crossover of around 150hz and no content from other speakers should be routed by the bass management system as was stated above is the low pass used my emotiva. Problem is it doesn’t. It’s directly linked to the center channel speaker regardless of LFE setting. If I set mains to large to get the LFE setting for the center sub I can change the output on the LFE directly proportionate to the crossover used on the center channel which shouldn’t be the case. Now on the other side of things if I select mono on the center sub I would expect all content below any speakers crossover to be diverted to the sub via bass management system. For this if I send an LFE signal to the XMC it should play 0-150hz if I have the crossover set to 200hz on the center and run a sweep on the center speaker from 10-1000 I would expect information sent to the sub from 10-200 and speaker above 200hz (not exactly as that’s not how filters work but for explanation sake I simplified). This should give me a sub reading from 0-200hz but it doesn’t For my clarification, what do you mean by sending an LFE signal. You should be just sending a sweep. One not encoded. Just as you would measure a full range speaker. If you are somehow encoding the signal so it is decoded as lfe... This may be your issue. |
|
|
|
Post by foggy1956 on May 3, 2020 12:22:42 GMT -5
Got the 2x4 advanced which gives me enough time. I may purchase a minidsp with Dirac capability dependent on how things work out. Only use the subs for movies, music is happiest in reference stereo to my ears. The 2x4HD can be converted to Dirac 2 for $199 and a firmware upgrade. Thank you for the info, does Dirac 2 incorporate the new bass algorithm that you are aware of? |
|
|
|
Post by kahodges1721 on May 3, 2020 12:24:28 GMT -5
You are correct LFE and Mono are very different which is my point. LFE should be independent of any crossover and holds content up to 120hz as you stated. In most receivers you can carry this channel low pass filter to as high as 250hz. Unless you are applying filters to the sub channel this doesn’t really matter since setting it above 120 provided no difference (it’s basically just applying double filters since the master is encoded with a filter of 120).This changes though when you run REW you can send any signal over LFE. Running LFE on the XMC center sub should show a crossover of around 150hz and no content from other speakers should be routed by the bass management system as was stated above is the low pass used my emotiva. Problem is it doesn’t. It’s directly linked to the center channel speaker regardless of LFE setting. If I set mains to large to get the LFE setting for the center sub I can change the output on the LFE directly proportionate to the crossover used on the center channel which shouldn’t be the case. Now on the other side of things if I select mono on the center sub I would expect all content below any speakers crossover to be diverted to the sub via bass management system. For this if I send an LFE signal to the XMC it should play 0-150hz if I have the crossover set to 200hz on the center and run a sweep on the center speaker from 10-1000 I would expect information sent to the sub from 10-200 and speaker above 200hz (not exactly as that’s not how filters work but for explanation sake I simplified). This should give me a sub reading from 0-200hz but it doesn’t For my clarification, what do you mean by sending an LFE signal. You should be just sending a sweep. One not encoded. Just as you would measure a full range speaker. If you are somehow encoding the signal so it is decoded as lfe... This may be your issue. That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover |
|
|
|
Post by megash0n on May 3, 2020 12:31:10 GMT -5
For my clarification, what do you mean by sending an LFE signal. You should be just sending a sweep. One not encoded. Just as you would measure a full range speaker. If you are somehow encoding the signal so it is decoded as lfe... This may be your issue. That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover I'm sure we are just crossing wires because we aren't in a room talking this over. But, an LFE encoded signal is not the same as the LFE channel. The LFE channel could be configured as lfe or mono. In my opinion, you aren't concerned whatsoever in an lfe encoded signal. That should do what it does regardless. When selecting the ASIO driver channel that is directly tied to the center sub, a basic pcm audio sweep from 20hz to 200hz or so should be played. Nothing encoded. If this is the case, and you are still getting a crossover applied, then something is broken or you are actually choosing the wrong channel. In my opinion. |
|
|
|
Post by kahodges1721 on May 3, 2020 12:49:08 GMT -5
That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover I'm sure we are just crossing wires because we aren't in a room talking this over. But, an LFE encoded signal is not the same as the LFE channel. The LFE channel could be configured as lfe or mono. In my opinion, you aren't concerned whatsoever in an lfe encoded signal. That should do what it does regardless. When selecting the ASIO driver channel that is directly tied to the center sub, a basic pcm audio sweep from 20hz to 200hz or so should be played. Nothing encoded. If this is the case, and you are still getting a crossover applied, then something is broken or you are actually choosing the wrong channel. In my opinion. Yeah would be much easier in the same room lol. LFE encoded signal sent to the XMC shouldn’t change the REW reading no matter if set to mono or LFE. Bc even in mono the sub still controls the LFE signal. Only difference is in mono it’s being asked to also handle anything below crossover points of speakers. Problem is that’s not how it works here. When set to mono an LFE only signal only plays to the crossover. So for instance let’s say any channel has a crossover of 80. If I then send an LFE only signal to the Xmc2 it applies a crossover at 80. Nothing else should be in play here at all. That’s how ASIO and REW work |
|
|
|
Post by megash0n on May 3, 2020 12:53:40 GMT -5
I'm sure we are just crossing wires because we aren't in a room talking this over. But, an LFE encoded signal is not the same as the LFE channel. The LFE channel could be configured as lfe or mono. In my opinion, you aren't concerned whatsoever in an lfe encoded signal. That should do what it does regardless. When selecting the ASIO driver channel that is directly tied to the center sub, a basic pcm audio sweep from 20hz to 200hz or so should be played. Nothing encoded. If this is the case, and you are still getting a crossover applied, then something is broken or you are actually choosing the wrong channel. In my opinion. Yeah would be much easier in the same room lol. LFE encoded signal sent to the XMC shouldn’t change the REW reading no matter if set to mono or LFE. Bc even in mono the sub still controls the LFE signal. Only difference is in mono it’s being asked to also handle anything below crossover points of speakers. Problem is that’s not how it works here. When set to mono an LFE only signal only plays to the crossover. So for instance let’s say any channel has a crossover of 80. If I then send an LFE only signal to the Xmc2 it applies a crossover at 80. Nothing else should be in play here at all. That’s how ASIO and REW work I think I'm following where your potentially "stuck". You are making the assumption I think that this LFE signal is actually encoded as LFE. Which, I would agree with you in logic. But, from a routing perspective, LFE wouldn't be routed to any other channel besides designated LFE channels. Is it possible that you are expecting this to be an lfe encoded signal when it is nothing more than a pcm audio sweep with no encoding at all? Think about that for a minute. I keep trying to take a step back and rework thru your steps to make sure I'm not missing some detail or information. Take a step back, minimize this down to the lowest number of variables and go from there. I'm not aware of REW being able to send encoded signals. Doesn't mean it can't. |
|
|
|
Post by kahodges1721 on May 3, 2020 12:56:56 GMT -5
Yeah would be much easier in the same room lol. LFE encoded signal sent to the XMC shouldn’t change the REW reading no matter if set to mono or LFE. Bc even in mono the sub still controls the LFE signal. Only difference is in mono it’s being asked to also handle anything below crossover points of speakers. Problem is that’s not how it works here. When set to mono an LFE only signal only plays to the crossover. So for instance let’s say any channel has a crossover of 80. If I then send an LFE only signal to the Xmc2 it applies a crossover at 80. Nothing else should be in play here at all. That’s how ASIO and REW work I think I'm following where your potentially "stuck". You are making the assumption I think that this LFE signal is actually encoded as LFE. Which, I would agree with you in logic. But, from a routing perspective, LFE wouldn't be routed to any other channel besides designated LFE channels. Is it possible that you are expecting this to be an lfe encoded signal when it is nothing more than a pcm audio sweep with no encoding at all? Think about that for a minute. I keep trying to take a step back and rework thru your steps to make sure I'm not missing some detail or information. Take a step back, minimize this down to the lowest number of variables and go from there. I'm not aware of REW being able to send encoded signals. Doesn't mean it can't. Yes REW can send encoded channels. When you set windows to 7.1 or whatever the receiver needs to know where to send the signal. So when you select HDMI1 it’s encoded for the left front hdmi2 is encoded for the left front hdmi3 for the center and hdmi4 for LFE. If it wasn’t this way then when you ran any sweep it would output from all the channels. |
|
KeithL
Administrator 
Posts: 10,273
|
Post by KeithL on May 3, 2020 13:05:49 GMT -5
Here is an important note on using REW most effectively.... IN THE CONTEXT OF ROOM MEASUREMENT AND CORRECTION... The most significant factors influencing the way things sound are your speakers, the room itself, and the location of your speakers in the room. In this context, you can assume that any decent solid state amplifier will be perfectly flat, and that all of its channels will be the same. Likewise, any preamp or processor, with bass management disabled, can likewise be considered to be perfectly flat.
(It's also worth noting that, when you use the "XMC-1 setting" on REW, REW treats each channel as entirely separate, and doesn't "know" which channel it is measuring.) Therefore, rather than mess with a computer, and attempt to control which channel you're sending the test signal to... The easiest solution is to use one signal output channel for all of your measurements and simply "walk the interconnect across the amplifier inputs". (And, strictly speaking, the most accurate way to measure multiple channels is to use a single signal source for all of your measurements anyway.)
In other words: - Unplug all of the interconnects from all of the inputs on your power amplifier(s). - Choose ONE OUTPUT that is going to be your "signal generator" (let's say LEFT FRONT).
- Arrange for your test signal to play from the LEFT FRONT output on your processor (or simply use a cable directly from the sound card on your computer). (This cable is now your "signal generator source" - which will deliver exactly the same test signal to each channel on your amplifier.) - Now, connect that cable to the input of the amplifier that goes to your Right Front speaker... and take your measurements for that speaker... (If you're having REW calculate corrections, and using the XMC-1 setting, you can tell REW to name that output file properly for "XMC-1 Right Front").
- Now, move that cable to the input on the amplifier that goes to your Left Front speaker and measure that.... - Repeat until you've measured all of your channels. Alternately, you can use a stereo output from the computer, and set the processor to REFERENCE STEREO / ALL STEREO. This will allow you to measure everything including the EQ adjustments already applied to each channel in the processor. Or set it to just plain STEREO and measure everything INCLUDING the bass management and EQ filters already active on each channel. If it's easier, you could even use one channel on the amplifier too, and connect each speaker to that speaker output, one after the other.
The "take away" from this is that sometimes it is much simpler to move wires than to try and figure out complex switching options.
Here's a link to a short paper on the subject of using REW with the RMC-1, RMC-1L, and XMC-2... which touches on this issue (in this context it also applies to the XMC-1):
For my clarification, what do you mean by sending an LFE signal. You should be just sending a sweep. One not encoded. Just as you would measure a full range speaker. If you are somehow encoding the signal so it is decoded as lfe... This may be your issue. That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover |
|
|
|
Post by kahodges1721 on May 3, 2020 13:12:47 GMT -5
Here is an important note on using REW most effectively.... IN THE CONTEXT OF ROOM MEASUREMENT AND CORRECTION... The most significant factors influencing the way things sound are your speakers, the room itself, and the location of your speakers in the room. In this context, you can assume that any decent solid state amplifier will be perfectly flat, and that all of its channels will be the same. Likewise, any preamp or processor, with bass management disabled, can likewise be considered to be perfectly flat.
(It's also worth noting that, when you use the "XMC-1 setting" on REW, REW treats each channel as entirely separate, and doesn't "know" which channel it is measuring.) Therefore, rather than mess with a computer, and attempt to control which channel you're sending the test signal to... The easiest solution is to use one signal output channel for all of your measurements and simply "walk the interconnect across the amplifier inputs". (And, strictly speaking, the most accurate way to measure multiple channels is to use a single signal source for all of your measurements anyway.)
In other words: - Unplug all of the interconnects from all of the inputs on your power amplifier(s). - Choose ONE OUTPUT that is going to be your "signal generator" (let's say LEFT FRONT).
- Arrange for your test signal to play from the LEFT FRONT output on your processor (or simply use a cable directly from the sound card on your computer). (This cable is now your "signal generator source" - which will deliver exactly the same test signal to each channel on your amplifier.) - Now, connect that cable to the input of the amplifier that goes to your Right Front speaker... and take your measurements for that speaker... (If you're having REW calculate corrections, and using the XMC-1 setting, you can tell REW to name that output file properly for "XMC-1 Right Front").
- Now, move that cable to the input on the amplifier that goes to your Left Front speaker and measure that.... - Repeat until you've measured all of your channels. Alternately, you can use a stereo output from the computer, and set the processor to REFERENCE STEREO / ALL STEREO. This will allow you to measure everything including the EQ adjustments already applied to each channel in the processor. Or set it to just plain STEREO and measure everything INCLUDING the bass management and EQ filters already active on each channel. If it's easier, you could even use one channel on the amplifier too, and connect each speaker to that speaker output, one after the other.
The "take away" from this is that sometimes it is much simpler to move wires than to try and figure out complex switching options.
Here's a link to a short paper on the subject of using REW with the RMC-1, RMC-1L, and XMC-2... which touches on this issue (in this context it also applies to the XMC-1):
That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover Hey Keith thanks for chiming in. I do like your method and will give it a read. How does that method allow for impulse measurements and such? Seems like it would make it very complicated to time align the speakers vs just use ASIO to select the right speaker as it is meant to be used. To clarify I’m using these measurements to verify subwoofer design. Here is an overlay of the hornres and previously measured subwoofers from a different receiver.  This is what I am trying to recreate. Issue is it seems the LFE channel might have a bug. I will do some more digging and see if I can pinpoint exactly what is going on here and report it out to you. Something is Off for sure as I’ve done this so many times I can’t count but can’t get it to work here. |
|
|
|
Post by megash0n on May 3, 2020 13:16:42 GMT -5
I think I'm following where your potentially "stuck". You are making the assumption I think that this LFE signal is actually encoded as LFE. Which, I would agree with you in logic. But, from a routing perspective, LFE wouldn't be routed to any other channel besides designated LFE channels. Is it possible that you are expecting this to be an lfe encoded signal when it is nothing more than a pcm audio sweep with no encoding at all? Think about that for a minute. I keep trying to take a step back and rework thru your steps to make sure I'm not missing some detail or information. Take a step back, minimize this down to the lowest number of variables and go from there. I'm not aware of REW being able to send encoded signals. Doesn't mean it can't. Yes REW can send encoded channels. When you set windows to 7.1 or whatever the receiver needs to know where to send the signal. So when you select HDMI1 it’s encoded for the left front hdmi2 is encoded for the left front hdmi3 for the center and hdmi4 for LFE. If it wasn’t this way then when you ran any sweep it would output from all the channels. I would consider that "channel routing" and not encoding. This is an area I will admit I'm not very knowledgeable. I'm only trying to help you think thru this. Encoding, to me, is Dolby, DTS, pcm, etc. When you route your PCM signal to the sub(hmdi4) , it is still PCM encoded. If you were to be able to strip out the .1 track of a DTS stream, and "route" it to hdmi4, then I would expect to get the results you are looking for. My logic still doesn't explain why you have a crossover being applied when sweeping hdmi4 unless center sub is set for LFE or you are using the wrong output. It would make sense that 4 is the sub. Left, Center, Right, Sub(4). The ASIO drivers have been known to be buggy as well. To step all the way back, what was wrong with this approach: Set center sub as mono. Set left channel as small with the highest crossover point. Use the Java drivers in REW. Do a sweep using the left channel as a Normal speaker up to around 240hz. (I think there is an option to send a subwoofer signal instead, but that may just be for setting levels) This is your sub measurement. Now, go into minidsp and disable the sub output. Set the left channel to large and run a sweep from 20 - whatever. This is your left channel by itself sweep. Decide on crossover and adjust timing and phase in minidsp. Set left channel back to small and run sweep. Adjust as needed. |
|
|
|
Post by megash0n on May 3, 2020 13:21:50 GMT -5
Here is an important note on using REW most effectively.... IN THE CONTEXT OF ROOM MEASUREMENT AND CORRECTION... The most significant factors influencing the way things sound are your speakers, the room itself, and the location of your speakers in the room. In this context, you can assume that any decent solid state amplifier will be perfectly flat, and that all of its channels will be the same. Likewise, any preamp or processor, with bass management disabled, can likewise be considered to be perfectly flat.
(It's also worth noting that, when you use the "XMC-1 setting" on REW, REW treats each channel as entirely separate, and doesn't "know" which channel it is measuring.) Therefore, rather than mess with a computer, and attempt to control which channel you're sending the test signal to... The easiest solution is to use one signal output channel for all of your measurements and simply "walk the interconnect across the amplifier inputs". (And, strictly speaking, the most accurate way to measure multiple channels is to use a single signal source for all of your measurements anyway.)
In other words: - Unplug all of the interconnects from all of the inputs on your power amplifier(s). - Choose ONE OUTPUT that is going to be your "signal generator" (let's say LEFT FRONT).
- Arrange for your test signal to play from the LEFT FRONT output on your processor (or simply use a cable directly from the sound card on your computer). (This cable is now your "signal generator source" - which will deliver exactly the same test signal to each channel on your amplifier.) - Now, connect that cable to the input of the amplifier that goes to your Right Front speaker... and take your measurements for that speaker... (If you're having REW calculate corrections, and using the XMC-1 setting, you can tell REW to name that output file properly for "XMC-1 Right Front").
- Now, move that cable to the input on the amplifier that goes to your Left Front speaker and measure that.... - Repeat until you've measured all of your channels. Alternately, you can use a stereo output from the computer, and set the processor to REFERENCE STEREO / ALL STEREO. This will allow you to measure everything including the EQ adjustments already applied to each channel in the processor. Or set it to just plain STEREO and measure everything INCLUDING the bass management and EQ filters already active on each channel. If it's easier, you could even use one channel on the amplifier too, and connect each speaker to that speaker output, one after the other.
The "take away" from this is that sometimes it is much simpler to move wires than to try and figure out complex switching options.
Here's a link to a short paper on the subject of using REW with the RMC-1, RMC-1L, and XMC-2... which touches on this issue (in this context it also applies to the XMC-1):
That is how REW is configured. HDMI 4 will only output to the LFE channel. So when sending a signal across LFE it should show the low pass used by emotiva every time. If I swap to HDMI3 output which is configured for center channel in REW I would expect the center channel to send its content to the sub below crossover Pretty much what he said. The only thing I would add is that you may need edit your distance setting each time you measurement the next channel as crossover points and timing may be relevant. This is assuming we are integrating things obviously. The ASIO drivers do make all of this a lot easier as long as they work. I prefer this as opposed to the long way, but those drivers crash my computer often |
|
|
|
Post by megash0n on May 3, 2020 13:25:02 GMT -5
Here is an important note on using REW most effectively.... IN THE CONTEXT OF ROOM MEASUREMENT AND CORRECTION... The most significant factors influencing the way things sound are your speakers, the room itself, and the location of your speakers in the room. In this context, you can assume that any decent solid state amplifier will be perfectly flat, and that all of its channels will be the same. Likewise, any preamp or processor, with bass management disabled, can likewise be considered to be perfectly flat.
(It's also worth noting that, when you use the "XMC-1 setting" on REW, REW treats each channel as entirely separate, and doesn't "know" which channel it is measuring.) Therefore, rather than mess with a computer, and attempt to control which channel you're sending the test signal to... The easiest solution is to use one signal output channel for all of your measurements and simply "walk the interconnect across the amplifier inputs". (And, strictly speaking, the most accurate way to measure multiple channels is to use a single signal source for all of your measurements anyway.)
In other words: - Unplug all of the interconnects from all of the inputs on your power amplifier(s). - Choose ONE OUTPUT that is going to be your "signal generator" (let's say LEFT FRONT).
- Arrange for your test signal to play from the LEFT FRONT output on your processor (or simply use a cable directly from the sound card on your computer). (This cable is now your "signal generator source" - which will deliver exactly the same test signal to each channel on your amplifier.) - Now, connect that cable to the input of the amplifier that goes to your Right Front speaker... and take your measurements for that speaker... (If you're having REW calculate corrections, and using the XMC-1 setting, you can tell REW to name that output file properly for "XMC-1 Right Front").
- Now, move that cable to the input on the amplifier that goes to your Left Front speaker and measure that.... - Repeat until you've measured all of your channels. Alternately, you can use a stereo output from the computer, and set the processor to REFERENCE STEREO / ALL STEREO. This will allow you to measure everything including the EQ adjustments already applied to each channel in the processor. Or set it to just plain STEREO and measure everything INCLUDING the bass management and EQ filters already active on each channel. If it's easier, you could even use one channel on the amplifier too, and connect each speaker to that speaker output, one after the other.
The "take away" from this is that sometimes it is much simpler to move wires than to try and figure out complex switching options.
Here's a link to a short paper on the subject of using REW with the RMC-1, RMC-1L, and XMC-2... which touches on this issue (in this context it also applies to the XMC-1):
Hey Keith thanks for chiming in. I do like your method and will give it a read. How does that method allow for impulse measurements and such? Seems like it would make it very complicated to time align the speakers vs just use ASIO to select the right speaker as it is meant to be used. To clarify I’m using these measurements to verify subwoofer design. Here is an overlay of the hornres and previously measured subwoofers from a different receiver. This is what I am trying to recreate. Issue is it seems the LFE channel might have a bug. I will do some more digging and see if I can pinpoint exactly what is going on here and report it out to you. Something is Off for sure as I’ve done this so many times I can’t count but can’t get it to work here. REW can be set to play a timing chime at the beginning of every sweep that I think you can use for this. This is one area ive tried to learn more about. |
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on May 3, 2020 13:46:48 GMT -5
That's an interesting question... and one with many nuances to it.
First off, as far as I recall, REW always measures one channel at a time... This is great for frequency response but more or less useless for combined impulse response from multiple speakers...
Second, even with ASIO, I would not be inclined to trust a computer to deliver multiple channels with precise enough timing for accurate impulse measurements. There are simply too many variables between the software, the ASIO driver, and the output. Superimposing those graphs assumes that you absolutely trust them to start at exactly the same point in time. I would also note that, in a room, your results are only going to be valid at exactly one spot in the room anyway.
Another factor is that, in a processor, there are multiple paths to the subwoofer.
For example, bass from the LFE channel goes straight to the sub, although there is a default low-pass filter in there.
However, with the bass management for the other channels, there are filters involved, each of which WILL introduce some phase shift. And the distance setting then adds to that a fixed time delay.
(Therefore, technically speaking, there will be no single solution that will be absolutely perfect for both LFE and bass managed content...)
The way I see it you have two choices....
Try to tune the impulse response for JUST THE SUBS.... (assuming that, once you get all the subs to act like one ideal sub, you're going to trust the other stuff to be right).
Or try to do so for "the entire system".... (remembering that the adjustment that is perfect for bass managed content my not be equally perfect for LFE content and vice versa). (The FIRST option there would be my choice.... get the subs right and trust the system to handle the details beyond that.)
However, in either situation, I would not even bother to plot separate graphs, and then try to carefully "align" them. There are simply too many variables.... and too many unknowns... and too much opportunity for the theory to not match the practical results.
What I would do is to hook up all the subs... in whatever manner I will be using them...
And play whatever test signal I prefer through all of them at once... (For example, play a test album where someone goes Whack Whack Whack on a drum, or a nice series of impulse clicks....)
Then I would sit the microphone at my listening position... And look at the ACTUAL IMPULSE RESPONSE AT THE LISTENING POSITION WHEN I PLAY THE TEST SIGNAL....
And I would adjust the timing on the subs until it looks good (one peak, or least ringing, or whatever you consider to be the most important goal)....
Doing it this way effectively "eliminates all variables, both known and unknown, and allows you to actually adjust what you will be hearing".
Here is an important note on using REW most effectively.... IN THE CONTEXT OF ROOM MEASUREMENT AND CORRECTION... The most significant factors influencing the way things sound are your speakers, the room itself, and the location of your speakers in the room. In this context, you can assume that any decent solid state amplifier will be perfectly flat, and that all of its channels will be the same. Likewise, any preamp or processor, with bass management disabled, can likewise be considered to be perfectly flat.
(It's also worth noting that, when you use the "XMC-1 setting" on REW, REW treats each channel as entirely separate, and doesn't "know" which channel it is measuring.) Therefore, rather than mess with a computer, and attempt to control which channel you're sending the test signal to... The easiest solution is to use one signal output channel for all of your measurements and simply "walk the interconnect across the amplifier inputs". (And, strictly speaking, the most accurate way to measure multiple channels is to use a single signal source for all of your measurements anyway.)
In other words: - Unplug all of the interconnects from all of the inputs on your power amplifier(s). - Choose ONE OUTPUT that is going to be your "signal generator" (let's say LEFT FRONT).
- Arrange for your test signal to play from the LEFT FRONT output on your processor (or simply use a cable directly from the sound card on your computer). (This cable is now your "signal generator source" - which will deliver exactly the same test signal to each channel on your amplifier.) - Now, connect that cable to the input of the amplifier that goes to your Right Front speaker... and take your measurements for that speaker... (If you're having REW calculate corrections, and using the XMC-1 setting, you can tell REW to name that output file properly for "XMC-1 Right Front").
- Now, move that cable to the input on the amplifier that goes to your Left Front speaker and measure that.... - Repeat until you've measured all of your channels. Alternately, you can use a stereo output from the computer, and set the processor to REFERENCE STEREO / ALL STEREO. This will allow you to measure everything including the EQ adjustments already applied to each channel in the processor. Or set it to just plain STEREO and measure everything INCLUDING the bass management and EQ filters already active on each channel. If it's easier, you could even use one channel on the amplifier too, and connect each speaker to that speaker output, one after the other.
The "take away" from this is that sometimes it is much simpler to move wires than to try and figure out complex switching options.
Here's a link to a short paper on the subject of using REW with the RMC-1, RMC-1L, and XMC-2... which touches on this issue (in this context it also applies to the XMC-1):
Hey Keith thanks for chiming in. I do like your method and will give it a read. How does that method allow for impulse measurements and such? Seems like it would make it very complicated to time align the speakers vs just use ASIO to select the right speaker as it is meant to be used. To clarify I’m using these measurements to verify subwoofer design. Here is an overlay of the hornres and previously measured subwoofers from a different receiver. This is what I am trying to recreate. Issue is it seems the LFE channel might have a bug. I will do some more digging and see if I can pinpoint exactly what is going on here and report it out to you. Something is Off for sure as I’ve done this so many times I can’t count but can’t get it to work here. |
|
