ttocs
Global Moderator 
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,115
|
Post by ttocs on May 4, 2021 1:02:42 GMT -5
Firstly, even though Emotiva may say it doesn't matter that the Emotiva Dirac Network Interface Box (EDNIB) is connected while attempting to attain an IP address for the processor, it does. When the EDNIB is connected it will interfere with the processor's ability to attain an IP so disconnect it, then boot the processor from cold, then connect the EDNIB after the processor has gotten its IP.
Secondly, I use a Mac Mini for my Dirac runs and keep the OSD at its minimum opacity. Not great, but it will work good enough.
I keep the mic at 100%, then lower the Master Volume down to around 30% or so and cycle through all of the speakers to get an initial response with the pink noise. Once all the speakers have sounded and show a level in their respective meters I choose the lowest one and Set it to such that it sounds at about -25dB or so and stop the noise. Then, I adjust all the other sliders to match close enough without them playing the pink noise, it's not necessary to play the pink noise for each one as the sliders are now pretty accurate after that initial response at the beginning of this paragraph. When it comes to the subwoofers I have chosen to have those play a few dB higher, at about -22dB. All this will make the measurement sweeps play loud enough that I plug my ears but not so loud as to be really annoying, somewhere around 88dB or so. YMMV
If your Master Volume does nothing, then there's something not right with the setup of the mic. If you're using a PC then I'm no help with that, but I know some PC setups need to be set or adjusted such that Dirac will play nice with the mic. I have the EMM-1 and two UMIK-1 mics and all are no trouble to use with Dirac and Mac OS.
Ditch the Emotiva supplied network switch if you can. It hinders more than helps. I've got the EDNIB and processor connected to a wifi extender directly because it runs more smoothly. It runs best if I connect them to the wifi router itself, but even so, it's more convenient to use this extender with my house arranged the way it is so I muster through with having to disconnect the EDNIB every now and then when I need to reboot the processor, otherwise, it's connected all the time.
|
|
KeithL
Administrator
Posts: 10,247
|
Post by KeithL on May 4, 2021 8:56:44 GMT -5
The processor and the Network Interface Box are entirely separate devices on your network (until the processor connects to the EDNIB). Therefore there is no particular reason why they should interfere with each other when they aren't actively connected to each other. (We normally leave ours running.)
However both get their IP addresses from your DHCP server... so it's possible for that to have some sort of odd issues.
And, of course, to some degree, everything on a network affects everything else on the network.
Networks are actually quite complex (although on home networks most of that complexity is hidden from the end user).
However there are two things you need to know about the EDNIB....
- First, if you shut the EDNIB down (using the Shutdown command on the processor), it will not wake up again until you reboot it by power cycling it.
- Second, and related, if the EDNIB is on, and your AC power flickers at some point, the EDNIB may restart improperly. If that happens the EDNIB won't work properly until you reboot it by power cycling it. (This generally won't damage anything. But, if you leave the EDNIB on all the time, this could easily happen during the night and go unnoticed.)
We have also been told that the Dirac Live software does not work properly with networks that use a NetMask other than 255.255.255.0 . 255.255.255.0 is the most common NetMask used by home networks. However, some mesh networks and network extenders use others, and Dirac Live may not work with some of them.
Firstly, even though Emotiva may say it doesn't matter that the Emotiva Dirac Network Interface Box (EDNIB) is connected while attempting to attain an IP address for the processor, it does. When the EDNIB is connected it will interfere with the processor's ability to attain an IP so disconnect it, then boot the processor from cold, then connect the EDNIB after the processor has gotten its IP. Secondly, I use a Mac Mini for my Dirac runs and keep the OSD at its minimum opacity. Not great, but it will work good enough. I keep the mic at 100%, then lower the Master Volume down to around 30% or so and cycle through all of the speakers to get an initial response with the pink noise. Once all the speakers have sounded and show a level in their respective meters I choose the lowest one and Set it to such that it sounds at about -25dB or so and stop the noise. Then, I adjust all the other sliders to match close enough without them playing the pink noise, it's not necessary to play the pink noise for each one as the sliders are now pretty accurate after that initial response at the beginning of this paragraph. When it comes to the subwoofers I have chosen to have those play a few dB higher, at about -22dB. All this will make the measurement sweeps play loud enough that I plug my ears but not so loud as to be really annoying, somewhere around 88dB or so. YMMV If your Master Volume does nothing, then there's something not right with the setup of the mic. If you're using a PC then I'm no help with that, but I know some PC setups need to be set or adjusted such that Dirac will play nice with the mic. I have the EMM-1 and two UMIK-1 mics and all are no trouble to use with Dirac and Mac OS. Ditch the Emotiva supplied network switch if you can. It hinders more than helps. I've got the EDNIB and processor connected to a wifi extender directly because it runs more smoothly. It runs best if I connect them to the wifi router itself, but even so, it's more convenient to use this extender with my house arranged the way it is so I muster through with having to disconnect the EDNIB every now and then when I need to reboot the processor, otherwise, it's connected all the time. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,115
|
Post by ttocs on May 4, 2021 11:40:30 GMT -5
The processor and the Network Interface Box are entirely separate devices on your network (until the processor connects to the EDNIB). Therefore there is no particular reason why they should interfere with each other when they aren't actively connected to each other. (We normally leave ours running.)
However both get their IP addresses from your DHCP server... so it's possible for that to have some sort of odd issues.
And, of course, to some degree, everything on a network affects everything else on the network.
Networks are actually quite complex (although on home networks most of that complexity is hidden from the end user).
However there are two things you need to know about the EDNIB....
- First, if you shut the EDNIB down (using the Shutdown command on the processor), it will not wake up again until you reboot it by power cycling it.
- Second, and related, if the EDNIB is on, and your AC power flickers at some point, the EDNIB may restart improperly. If that happens the EDNIB won't work properly until you reboot it by power cycling it. (This generally won't damage anything. But, if you leave the EDNIB on all the time, this could easily happen during the night and go unnoticed.)
We have also been told that the Dirac Live software does not work properly with networks that use a NetMask other than 255.255.255.0 . 255.255.255.0 is the most common NetMask used by home networks. However, some mesh networks and network extenders use others, and Dirac Live may not work with some of them.
Firstly, even though Emotiva may say it doesn't matter that the Emotiva Dirac Network Interface Box (EDNIB) is connected while attempting to attain an IP address for the processor, it does. When the EDNIB is connected it will interfere with the processor's ability to attain an IP so disconnect it, then boot the processor from cold, then connect the EDNIB after the processor has gotten its IP. Secondly, I use a Mac Mini for my Dirac runs and keep the OSD at its minimum opacity. Not great, but it will work good enough. I keep the mic at 100%, then lower the Master Volume down to around 30% or so and cycle through all of the speakers to get an initial response with the pink noise. Once all the speakers have sounded and show a level in their respective meters I choose the lowest one and Set it to such that it sounds at about -25dB or so and stop the noise. Then, I adjust all the other sliders to match close enough without them playing the pink noise, it's not necessary to play the pink noise for each one as the sliders are now pretty accurate after that initial response at the beginning of this paragraph. When it comes to the subwoofers I have chosen to have those play a few dB higher, at about -22dB. All this will make the measurement sweeps play loud enough that I plug my ears but not so loud as to be really annoying, somewhere around 88dB or so. YMMV If your Master Volume does nothing, then there's something not right with the setup of the mic. If you're using a PC then I'm no help with that, but I know some PC setups need to be set or adjusted such that Dirac will play nice with the mic. I have the EMM-1 and two UMIK-1 mics and all are no trouble to use with Dirac and Mac OS. Ditch the Emotiva supplied network switch if you can. It hinders more than helps. I've got the EDNIB and processor connected to a wifi extender directly because it runs more smoothly. It runs best if I connect them to the wifi router itself, but even so, it's more convenient to use this extender with my house arranged the way it is so I muster through with having to disconnect the EDNIB every now and then when I need to reboot the processor, otherwise, it's connected all the time. Thanks for the background info here. FYI I've tested the EDNIB operation with the Emo supplied switch (not good), a managed switch (the best of all items I tested and worked without any fails of any kind while the EDNIB was connected and powered 100% of the time with computer/EDNIB/XMC-2 wired to it), 2 different brand wifi routers (the Asus works very well with regard to the EDNIB not interfering with XMC-2 gaining an IP address, but not perfectly like the managed switch), and the wifi extender that is now being used (3rd best, but most convenient for my home network - all devices considered). |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 4, 2021 12:43:14 GMT -5
The processor and the Network Interface Box are entirely separate devices on your network (until the processor connects to the EDNIB). Therefore there is no particular reason why they should interfere with each other when they aren't actively connected to each other. (We normally leave ours running.)
However both get their IP addresses from your DHCP server... so it's possible for that to have some sort of odd issues.
And, of course, to some degree, everything on a network affects everything else on the network.
Networks are actually quite complex (although on home networks most of that complexity is hidden from the end user).
However there are two things you need to know about the EDNIB....
- First, if you shut the EDNIB down (using the Shutdown command on the processor), it will not wake up again until you reboot it by power cycling it.
- Second, and related, if the EDNIB is on, and your AC power flickers at some point, the EDNIB may restart improperly. If that happens the EDNIB won't work properly until you reboot it by power cycling it. (This generally won't damage anything. But, if you leave the EDNIB on all the time, this could easily happen during the night and go unnoticed.)
We have also been told that the Dirac Live software does not work properly with networks that use a NetMask other than 255.255.255.0 . 255.255.255.0 is the most common NetMask used by home networks. However, some mesh networks and network extenders use others, and Dirac Live may not work with some of them.
Thanks for the background info here. FYI I've tested the EDNIB operation with the Emo supplied switch (not good), a managed switch (the best of all items I tested and worked without any fails of any kind while the EDNIB was connected and powered 100% of the time with computer/EDNIB/XMC-2 wired to it), 2 different brand wifi routers (the Asus works very well with regard to the EDNIB not interfering with XMC-2 gaining an IP address, but not perfectly like the managed switch), and the wifi extender that is now being used (3rd best, but most convenient for my home network - all devices considered). Is it possible to assign static IPs from not only the router [DHCP] but also statically set the IPs in each device [Network Adapters]? I'd think doing so might eliminate any conflict in acquiring IP numbers [boot order etc]? Be curious as to the results if someone tries to statically assign IPs. Just remember not to use IPs ending in .0 or .255 as they're usually reserved for network protocols. I mean there are benefits to statically assigning IPs not only in speed but also security. The downside is that any new device that's welcome needs be manually assigned in the DCHP server's settings to connect to the network. |
|
KeithL
Administrator
Posts: 10,247
|
Post by KeithL on May 4, 2021 16:26:27 GMT -5
You most certainly can enter addresses into at least some devices manually - but unfortunately your supposition is exactly backwards... It requires careful consideration to decide what addresses will work with your equipment and your network...
(Much like phone numbers... network addresses are not at all random.)
The biggest issue to be aware of is the possibility of an address conflict - which is what happens when the same address is assigned to two devices. If you have an address conflict, either or both conflicting devices may act erratically, or may work at some times and not others. But, even worse, since the address is in fact assigned to an active device, that address will often respond PROPERLY to a Ping. (So, you Ping the address, and everything looks fine... although you may get odd results if you Ping it from different locations on the network.)
When you use DHCP the normal way the DHCP server assigns all the addresses. It also keeps careful track to ensure that it never assigns the same address to two devices. And it knows to use addresses that follow both the standards of Internet routing and the configuration of your network.
Some DHCP servers allow you to manually enter specific addresses that should always be used for specific devices.
However, this is still not fully manual, since the DHCP server is aware of those addresses, and so won't assign them to another device. (In most cases your DHCP server is only going to allow you to set addresses that will work with your network.)
HOWEVER, if you manually enter addresses directly into one or more devices, THE DHCP SERVER IS UNAWARE THAT YOU HAVE DONE SO. Therefore there is nothing to check the addresses you choose to make sure they're going to work.
This means that: - there is nothing to prevent you from assigning the same address to multiple devices - there is nothing to prevent the DHCP server from assigning the address you used to another device (the DHCP server doesn't know you assigned the address) - (And, even if you checked and confirmed that the address isn't being used at the moment, there's nothing to stop DHCP from assigning it to another device later) - and, finally, there's nothing to prevent you from assigning an address that isn't valid for your network Things get even more complicated if you have a more complicated network - like one that uses routers and network extenders. Your router and DHCP server are configured to work together - so your router "knows" how to route traffic to and from addresses assigned by your DHCP server.
And, when it comes to things that are connected to the Internet, there are all sorts of very specific "rules and regulations".
For example, if you have a router that is connected to the Internet, then you may ONLY use addresses starting with 192.168 , 172.16 , or 10.0 on your network. (Routers that are designed to connect to the Internet are pre-programmed to "know" that these addresses are reserved for private networks.)
And, if you violate these rules, really odd things may happen, or things may simply not work. It is quite possible to manually assign addresses successfully... But you have to know quite a bit about your network, and configure several things properly, or there's a good chance it won't work as you expect. Thanks for the background info here. FYI I've tested the EDNIB operation with the Emo supplied switch (not good), a managed switch (the best of all items I tested and worked without any fails of any kind while the EDNIB was connected and powered 100% of the time with computer/EDNIB/XMC-2 wired to it), 2 different brand wifi routers (the Asus works very well with regard to the EDNIB not interfering with XMC-2 gaining an IP address, but not perfectly like the managed switch), and the wifi extender that is now being used (3rd best, but most convenient for my home network - all devices considered). Is it possible to assign static IPs from not only the router [DHCP] but also statically set the IPs in each device [Network Adapters]? I'd think doing so might eliminate any conflict in acquiring IP numbers [boot order etc]? Be curious as to the results if someone tries to statically assign IPs. Just remember not to use IPs ending in .0 or .255 as they're usually reserved for network protocols. I mean there are benefits to statically assigning IPs not only in speed but also security. The downside is that any new device that's welcome needs be manually assigned in the DCHP server's settings to connect to the network. |
|
|
|
Post by autocrat on May 4, 2021 16:41:10 GMT -5
Best way, IMO, is to have DHCP active on the device but reserve an IP address on the router for that device.
|
|
KeithL
Administrator
Posts: 10,247
|
Post by KeithL on May 4, 2021 16:52:49 GMT -5
That is a good solution if you specifically want a certain device to have the same address every time. But that's about all it will accomplish.
If you do it that way, you are still relying on DHCP on the device to fetch the address from the DHCP server. (So, if there is an issue with DHCP, this isn't likely to solve it.)
If you are really convinced that the DHCP on your device is not working properly then here's a better way... Reserve an IP address for the device on your DHCP server. Then enter THE SAME ADDRESS into your device MANUALLY.
This way you get to enter the address into your device manually... thus avoiding any possible DHCP problems. But, since the DHCP server allowed you to reserve the address, you know that it's a valid address, and that the DHCP server knows about it.
ALSO, BE VERY VERY SURE THAT YOU DON'T HAVE TWO ROUTERS, WITH TWO SEPARATE DHCP SERVERS, HANDING OUT ADDRESSES. That way lies madness (or at least serious frustration).  Best way, IMO, is to have DHCP active on the device but reserve an IP address on the router for that device. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 4, 2021 16:54:13 GMT -5
You most certainly can enter addresses into at least some devices manually - but unfortunately your supposition is exactly backwards... It requires careful consideration to decide what addresses will work with your equipment and your network...
(Much like phone numbers... network addresses are not at all random.)
The biggest issue to be aware of is the possibility of an address conflict - which is what happens when the same address is assigned to two devices. If you have an address conflict, either or both conflicting devices may act erratically, or may work at some times and not others. But, even worse, since the address is in fact assigned to an active device, that address will often respond PROPERLY to a Ping. (So, you Ping the address, and everything looks fine... although you may get odd results if you Ping it from different locations on the network.)
When you use DHCP the normal way the DHCP server assigns all the addresses. It also keeps careful track to ensure that it never assigns the same address to two devices. And it knows to use addresses that follow both the standards of Internet routing and the configuration of your network.
Some DHCP servers allow you to manually enter specific addresses that should always be used for specific devices.
However, this is still not fully manual, since the DHCP server is aware of those addresses, and so won't assign them to another device. (In most cases your DHCP server is only going to allow you to set addresses that will work with your network.)
HOWEVER, if you manually enter addresses directly into one or more devices, THE DHCP SERVER IS UNAWARE THAT YOU HAVE DONE SO. Therefore there is nothing to check the addresses you choose to make sure they're going to work.
This means that: - there is nothing to prevent you from assigning the same address to multiple devices - there is nothing to prevent the DHCP server from assigning the address you used to another device (the DHCP server doesn't know you assigned the address) - (And, even if you checked and confirmed that the address isn't being used at the moment, there's nothing to stop DHCP from assigning it to another device later) - and, finally, there's nothing to prevent you from assigning an address that isn't valid for your network Things get even more complicated if you have a more complicated network - like one that uses routers and network extenders. Your router and DHCP server are configured to work together - so your router "knows" how to route traffic to and from addresses assigned by your DHCP server.
And, when it comes to things that are connected to the Internet, there are all sorts of very specific "rules and regulations".
For example, if you have a router that is connected to the Internet, then you may ONLY use addresses starting with 192.168 , 172.16 , or 10.0 on your network. (Routers that are designed to connect to the Internet are pre-programmed to "know" that these addresses are reserved for private networks.)
And, if you violate these rules, really odd things may happen, or things may simply not work. It is quite possible to manually assign addresses successfully... But you have to know quite a bit about your network, and configure several things properly, or there's a good chance it won't work as you expect. Is it possible to assign static IPs from not only the router [DHCP] but also statically set the IPs in each device [Network Adapters]? I'd think doing so might eliminate any conflict in acquiring IP numbers [boot order etc]? Be curious as to the results if someone tries to statically assign IPs. Just remember not to use IPs ending in .0 or .255 as they're usually reserved for network protocols. I mean there are benefits to statically assigning IPs not only in speed but also security. The downside is that any new device that's welcome needs be manually assigned in the DCHP server's settings to connect to the network. I don't understand how my supposition is completely backwards or how so dumb a router may exists now-a-days in a network. And, to correctly assign static IP numbers one doesn't just input an IP number in the network adapter's configuration settings but also needs supply additional information such as Mac and Hostname in the router when manually configuring DHCP about the network adapters in question. Correct me if wrong but it isn't as simple as assigning the same static IP address in a network adapter by typing in the desired IP number or inputting the same IP in two different network adapters resulting in the DHCP server assigning the same IP number to them. The network adapter has a Mac address and the DHCP server knows by manual input the Mac address of the adapter in question and performs cks between the DHCP server and network adapter? I mean, it is possible to "try" to assign two different network adapters the same IP number but both will have different Mac addresses unless one spoofs the Mac address in the network adapter for another. If a network adapter comes online in the network and the IP it is requesting by static assignment doesn't match the Mac address the router shouldn't assign the IP. If a spoofed network adapter came online in the network requesting the same IP number then yes I can see where your argument is relevant. What are the chances anyone here has more than one network adapter w/ identical Mac addresses unless deliberately spoofed? I'm sure if anyone here has a spoofed network adapter then this is all remedial knowledge. This however is going beyond simplicity of correctly statically assigning IP numbers.  |
|
|
|
Post by jim80z on May 6, 2021 17:30:46 GMT -5
Hi All, Hope you and yours are safe and well. Im having some difficulty with DIRAC on my RMC-1. Quick list of versions and gear: - RMC-1 with FW 2.2 - Using a Mac with latest OS and version of DIRAC Live 3 - Emotiva RMC-1 Mic with CAL file - The RMC-1 has manual speaker sizing, speaker distance and levels (master vol @-20db using 75db external test tones) set for all speakers. I have read these settings have no effect on DIRAC measurements and are ignored but wanted to give a complete description) - 9.1.6 setup. Multiple subs fed by same output and I have 8 atmos speakers with the middle two on each side fed by 1 top middle output each. Again just being thorough. (side Q, when are we getting the expansion modules?) - Im following this document mehlau.net/audio/dirac-live-2-mic-speaker-levels/ and also the Emotiva videos. - the room is pretty much isolated from outside noise so no noises to spike the readings and clip My problem is I can't get past the volume calibration step as the main left or right speakers clip. Im having to reduce the volume, the mic and/or speaker level too low and still its a problem. And if I go too low on the mike and master volume it causes issues with the other speakers reading well. Anyone else striking this problem? Any suggestions? Thx Jim |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,115
|
Post by ttocs on May 6, 2021 19:53:13 GMT -5
Hi All, Hope you and yours are safe and well. Im having some difficulty with DIRAC on my RMC-1. Quick list of versions and gear: - RMC-1 with FW 2.2 - Using a Mac with latest OS and version of DIRAC Live 3 - Emotiva RMC-1 Mic with CAL file - The RMC-1 has manual speaker sizing, speaker distance and levels (master vol @-20db using 75db external test tones) set for all speakers. I have read these settings have no effect on DIRAC measurements and are ignored but wanted to give a complete description) - 9.1.6 setup. Multiple subs fed by same output and I have 8 atmos speakers with the middle two on each side fed by 1 top middle output each. Again just being thorough. (side Q, when are we getting the expansion modules?) - Im following this document mehlau.net/audio/dirac-live-2-mic-speaker-levels/ and also the Emotiva videos. - the room is pretty much isolated from outside noise so no noises to spike the readings and clip My problem is I can't get past the volume calibration step as the main left or right speakers clip. Im having to reduce the volume, the mic and/or speaker level too low and still its a problem. And if I go too low on the mike and master volume it causes issues with the other speakers reading well. Anyone else striking this problem? Any suggestions? Thx Jim Speaker Distance is totally irrelevant with Dirac and won't even show up in the Menu when a Dirac filter is active. Set the Levels to 0.0dB for all speakers. Dirac sets its own levels internally and remains hidden. After running Dirac you can check the RMC Levels and verify that they are all close and adjust if needed. On my XMC-2 I adjusted a few channels less than a couple dB, but my subwoofer channels needed to be adjusted down to -6dB. 1. With the current version of Dirac I leave the mic at 100%. 2. I adjust the Master Volume down to around -43dB or so before adjusting the Speaker Sliders, this is so the next step isn't very loud. 3. Next I hit the Play button for each speaker and let it play for a couple seconds so the meter can stabilize and get a reading. This is done for every speaker quickly. 4. Then I look for the weakest channel, play the pink noise, and adjust the Master Volume to get the meter in the zone, I choose to set it to about -25dB, but your system, speakers, and amps may like something a little different. I'm now done playing pink noise. Dirac knows how loud each channel is and is so I don't need to play the noise anymore for every other channel. 5. Now I simply adjust every slider to match the first channel that has been set - without playing the pink noise, it's not needed. Now I go to "Proceed to select arrangement". edit: This entire process takes about two minutes. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 6, 2021 23:48:09 GMT -5
Hi All, Hope you and yours are safe and well. Im having some difficulty with DIRAC on my RMC-1. Quick list of versions and gear: - RMC-1 with FW 2.2 - Using a Mac with latest OS and version of DIRAC Live 3 - Emotiva RMC-1 Mic with CAL file - The RMC-1 has manual speaker sizing, speaker distance and levels (master vol @-20db using 75db external test tones) set for all speakers. I have read these settings have no effect on DIRAC measurements and are ignored but wanted to give a complete description) - 9.1.6 setup. Multiple subs fed by same output and I have 8 atmos speakers with the middle two on each side fed by 1 top middle output each. Again just being thorough. (side Q, when are we getting the expansion modules?) - Im following this document mehlau.net/audio/dirac-live-2-mic-speaker-levels/ and also the Emotiva videos. - the room is pretty much isolated from outside noise so no noises to spike the readings and clip My problem is I can't get past the volume calibration step as the main left or right speakers clip. Im having to reduce the volume, the mic and/or speaker level too low and still its a problem. And if I go too low on the mike and master volume it causes issues with the other speakers reading well. Anyone else striking this problem? Any suggestions? Thx Jim Speaker Distance is totally irrelevant with Dirac and won't even show up in the Menu when a Dirac filter is active. Set the Levels to 0.0dB for all speakers. Dirac sets its own levels internally and remains hidden. After running Dirac you can check the RMC Levels and verify that they are all close and adjust if needed. On my XMC-2 I adjusted a few channels less than a couple dB, but my subwoofer channels needed to be adjusted down to -6dB. 1. With the current version of Dirac I leave the mic at 100%. 2. I adjust the Master Volume down to around -43dB or so before adjusting the Speaker Sliders, this is so the next step isn't very loud. 3. Next I hit the Play button for each speaker and let it play for a couple seconds so the meter can stabilize and get a reading. This is done for every speaker quickly. 4. Then I look for the weakest channel, play the pink noise, and adjust the Master Volume to get the meter in the zone, I choose to set it to about -25dB, but your system, speakers, and amps may like something a little different. I'm now done playing pink noise. Dirac knows how loud each channel is and is so I don't need to play the noise anymore for every other channel. 5. Now I simply adjust every slider to match the first channel that has been set - without playing the pink noise, it's not needed. Now I go to "Proceed to select arrangement". edit: This entire process takes about two minutes. "Speaker Distance is totally irrelevant with Dirac and won't even show up in the Menu when a Dirac filter is active." Makes sense, I mean how would Dirac compensate for speaker distance when measuring from multiple points? |
|
|
|
Post by leonski on May 6, 2021 23:59:31 GMT -5
In setup you MAY choose one postition as a reference. Software puts a pulse to each channel (somehow also defined in software) and measures time delay.
Time is VERY relevant to phase coherence. And if far enough apart, will even cancel...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 7, 2021 0:06:51 GMT -5
In setup you MAY choose one postition as a reference. Software puts a pulse to each channel (somehow also defined in software) and measures time delay. Time is VERY relevant to phase coherence. And if far enough apart, will even cancel... Yup, but I have no need to focus in Dirac on a single point measurement but received the best results from 13+ point measurements around a 3 wide Main Listening Position. Minidsp has a plugin where time alignment can be manually input:  Also, there's a Dirac Tab in the Minidsp plugin which shows the delays Dirac is using. Of course, delays can be calculated out in distance. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,115
|
Post by ttocs on May 7, 2021 8:12:19 GMT -5
"Speaker Distance is totally irrelevant with Dirac and won't even show up in the Menu when a Dirac filter is active." Makes sense, I mean how would Dirac compensate for speaker distance when measuring from multiple points? The timing measurement is done at the first mic location in the MLP location. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 7, 2021 8:36:47 GMT -5
"Speaker Distance is totally irrelevant with Dirac and won't even show up in the Menu when a Dirac filter is active." Makes sense, I mean how would Dirac compensate for speaker distance when measuring from multiple points? The timing measurement is done at the first mic location in the MLP location. Thanks, I take it that the first measuring point is pretty important to get centered then... If I remember right you're using both Emotiva and Minidsp? If so was just curious if you've noted the timing in the Minidsp's Plugin Dirac tab? Seems that various crossovers and actual processor latency shows up in Dirac time alignments. For example, the DDRC-88A has 13ms of latency in of itself. I was curious as to whether this is an issue when mixing multiple processors or running some speakers using Dirac and not others? Reason I ask is I may reintroduce Dirac again in my audio chain. |
|
|
|
Post by cygnusx on May 7, 2021 10:59:35 GMT -5
hopefully a simple question where do i place the mic when the couch has high backrests ? i got xmc2 7.1.6 and a reclining HT couch with high backrest (so high they support my head), the couch spans almost the width of the room, and i want the two middle positions to be at its very best the Dirac pictures shows a couch with low back, and to place the mic behind the couch. but with high backrests that doesnt seem right i have calibrated around 20 times, and each time the result i horrible (muddy, and emphasis on the right front instead of even stereo image), so i have gone back to tape measure and db meter and just set it up like that. i actually dragged in the living room pioneer (503) amp used its preouts and ran its simple calibration, and result was ok. when i calibrate i recline the couch to where it would be if we where watching a movie im aware that side surrounds (dipole) might have to be moved around picture sucks sun is blowing the picture out....rear couch is not being used (think of it as an acoustical treatment LOL)  ![]()  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 7, 2021 11:37:17 GMT -5
hopefully a simple question where do i place the mic when the couch has high backrests ? i got xmc2 7.1.6 and a reclining HT couch with high backrest (so high they support my head), the couch spans almost the width of the room, and i want the two middle positions to be at its very best the Dirac pictures shows a couch with low back, and to place the mic behind the couch. but with high backrests that doesnt seem right i have calibrated around 20 times, and each time the result i horrible (muddy, and emphasis on the right front instead of even stereo image), so i have gone back to tape measure and db meter and just set it up like that. i actually dragged in the living room pioneer (503) amp used its preouts and ran its simple calibration, and result was ok. when i calibrate i recline the couch to where it would be if we where watching a movie im aware that side surrounds (dipole) might have to be moved around picture sucks sun is blowing the picture out....rear couch is not being used (think of it as an acoustical treatment LOL) ![]() View AttachmentYou can skip the rear two chest level measurements which are obstructed by the high back of the couch thereby making an 11 point measurement. Or peek the tip of the mic over the back of the couch for those chest levels measurements for a 13 pointer. I'd skip any measurement which obstructs the microphone. I'd also recommend draping a blanket over the rear back of the couch to cut down on reflections while measuring. You could recline back the couch to your desired listening position and then make the measurements in relation to the persons in the Dirac model. For example, measurement 1 center of forehead of centered position while reclined back. I'd be interested in knowing whether this makes a lot of difference. I think that Dirac is creating a "bubble" in a multipoint measurement and as long as you're positioned in the bubble the variation should be minimized as to what you're hearing. Feel free to correct if wrong. |
|
|
|
Post by oleops on May 7, 2021 12:22:47 GMT -5
Just ran my first Dirac...
Thanks to AudioHTIT , links comes in handy..
And as a newbee I did not zeroed the speaker settings before the run.
So when I go into the meny I can still see my level settings.
I belive this settings are in adition to the dirac, as dirac only do the eq and dont bother with the settings?
But they should be 0 right? (to be even and only be raised or lowered to my likings... )
Can I zero them out now or must I run the dirac again, after?
Cannot run Dirac for a while now, must whait for a quiet home not happening that often :-p
I have not been listening so much after the Dirac run but i Think the backs where somewhat quiet...
Can also only be in my head as I know them to be 5dB down in my settings
Sorry for bad English.
Thanks!
|
|
|
|
Post by cygnusx on May 7, 2021 12:30:34 GMT -5
draping a blanket, had not thought about that, but it makes sense the leather is reflective...
so simply skip what would be obstructed
|
|
|
|
Post by AudioHTIT on May 7, 2021 12:38:03 GMT -5
|
|
