Configuring a PC as a 8-ch pre/pro experiment

[Vasr]

After several months of experimentation (mostly trying to understand poorly written documentation for multiple tools), I was able to set up a HTPC as a sort of AVR for multi-channel content with the following configuration

Using Voicemeeter, Equalizer APO and VSTHost.

It is really not a satisfactory replacement for an AVR because you cannot have a convenient IR remote for everything, content with DRM can be a problem and the new codecs with more than 8-ch are out of the scope. But otherwise works very well from 2.0 to 7.1, integrating with your favorite media player that just works as long as it plays to a selected/default sound device sending PCM and the DSP settings/processing apply to all of them. Crossover, delay and volume can be set for each channel independently to align phase/balance. There are no latency issues for lip sync that I have encountered.

I have decided to stay with this until a decent pre/pro unit comes into market at a sensible price. Right now, they are either too expensive with features I don't need or have bare minimum features + lots of glitches and need multiple external boxes to get the full function as above. An older AVR is also an alternate solution if it has pre-outs but you would have to buy an expensive DIrac hardware box or do more than half of the work above to get it to work in the PC for multi-channel sound.

With REW based EQ I was able to get sounds very good almost as good as ARC but Dirac has some magic that makes it noticeably better. Setting up Dirac was a pain though.

Until Dirac bass management is available for PC, the bass management here works fine to integrate a sub.

Maximum 8 channels total though for the PCM output. So you can set it up for 5.2 but not 7.2.

I was going to write up some notes on how to do this but it can be quite an undertaking to describe it in a fool-proof way. I am not sure how much interest there is in replicating such a set up here to make it worthwhile.

 

[Kal Rubinson]

There is interest. My interest in this is for music only (not video) but, AFAIK, DL3 bass management is available for PC as a beta now and there are ways to get more than 8 channels, albeit at a cost. As for DRM, one way around it is to rip the media.

 

[chaking]

I'm pretty curious to hear what you came up with. What hardware components did you use? Did you get (e)ARC working so you can have audio from components connected to the TV?

 

[Vasr]

There is interest. My interest in this is for music only (not video) but, AFAIK, DL3 bass management is available for PC as a beta now and there are ways to get more than 8 channels, albeit at a cost. As for DRM, one way around it is to rip the media.

While not a fan of multi-channel music, I have tried out a few multi-channel DSDs using foobar DSD plugin to test the set up. Works very well with room eq on all channels especially with Dirac.

I believe there was a stand-alone DL bass management that was beta tested to make it available for use with third party external boxes. There does not appear to be a DLP available yet with bass management incorporated for use with multi-channel DL 3.0 even in beta. I am supposed to be on their beta testers list but have not received any email about it yet.

 

[Vasr]

I'm pretty curious to hear what you came up with. What hardware components did you use? Did you get (e)ARC working so you can have audio from components connected to the TV?

The hardware is just a DIY PC with an old i7 (it is hardly under any load for any of this so I suspect even a current generation i3 will work). Had no problem software decoding DSD, etc.

Getting external multi-channel into the PC is an unsolved challenge. There are are no HDMI receiver cards available. So (e)ARC is not possible at the moment. The best best is to separate out the HDMI processing to an external box (you can get $50 HDMI switch/audio extractors) and tap the L and R PCM audio from it via optical. Getting digital multi-channel out of them is the problem yet to be solved due to licensing issues.

Right now my solution is to attach the Optical out of the TV (most TVs have them for sound bar use) into a sound card with Optical In. While higher end SoundBlaster cards have optical in ports, there is a cheaper Cmedia chip based card sold under StarTech and Syba labels for under $50. Works out of the box with no driver installs needed. They are also low profile cards and so can be used in half height HTPC boxes if needed.

It has a daughter card for analog line out above for testing but you can replace it with another main card as long as you have PCIe slots available to get multiple optical sources in.

The limitation is 2 channel PCM only through the optical in. No multi-channel bitstreams.

AC3 encodings are not recognized as such from the stream coming into the optical to do on-the-fly ac3 decoding inside the box (Windows/Industry conspiracy). Even if you were able to do that, the ac3 decoding in software for audio coming from a video content being played externally will be too slow to lipsync with what is already playing.

Video players rendering audio and video from inside the PC itself don't have this problem as they automatically delay the video decoding (in hardware) for the audio soft decoding to be in sync. So you can do any kind of supported audio codec for media being played from the PC without problems.

 

[Vasr]

I should also mention that this is a useful set up for use with Okto 8Pro and your choice of a multi-channel amp downstream.

The PC aggregates local content, streaming and optical sources into the USB out to the Okto and provides the common DSP features on all of them, Okto has the physical volume control. So you don't need anything else in the middle to do both music and HT (with the caveats mentioned above).

 

[Kal Rubinson]

I believe there was a stand-alone DL bass management that was beta tested to make it available for use with third party external boxes. There does not appear to be a DLP available yet with bass management incorporated for use with multi-channel DL 3.0 even in beta. I am supposed to be on their beta testers list but have not received any email about it yet.

You might check out the relevant threads on AVS Forum.

I should also mention that this is a useful set up for use with Okto 8Pro and your choice of a multi-channel amp downstream.

You can link/synch multiple OKTO DAC8 Pros for more channels.

 

[theshade]

After several months of experimentation (mostly trying to understand poorly written documentation for multiple tools), I was able to set up a HTPC as a sort of AVR for multi-channel content with the following configuration

I was going to write up some notes on how to do this but it can be quite an undertaking to describe it in a fool-proof way. I am not sure how much interest there is in replicating such a set up here to make it worthwhile.

Can you write up some notes? If people can easily do this then I guess the interest would be monumental even just due to the savings of not buying an avr or a prepro and amp.

 

[Vasr]

I will try to create some high level notes this week that will act as a rough guide but not a step-by-step instruction manual.

It can be described to a level of detail so that even those that are not tech savvy can easily do this but that takes a lot more effort than I can expend.

Note that it does have limitations which would make it a replacement for an AVR in only some specific use cases but not all.

 

[Vasr]

I am going to post some simple notes of the steps below broken up into multiple posts so they can be individually discussed rather than everything in one long post. So, please bear with me until I am done. I will break this up into the following posts:

1. Prepping the PC and requirements
2. Setting up Voicemeeter Banana as the core sound engine
3. Setting up Equalizer APO to work with Voicemeeter
4. Setting up speaker balance and delays
5. Setting up bass management
6. Adding REW generated room-eq filters
7. Adding multi-channel Dirac Live 3.x as an alternative
8. Adding an external source to the PC via optical input
9. Miscellaneous Topics

Pleased don't ask me for basic Windows concepts for Audio or details of how the above tools work. You can pick them up by googling. I am just going to list what you need to do with these tools to set it up.

 

[Vasr]

1. Prepping the PC and requirements
This set up is specifically to ensure multi-channel audio can be handled. So you need at least one output audio device on the PC capable of putting out multi-channel audio (at least 2.1 for any of these steps to be meaningful).

The simplest is if you have a 5.1 or 7.1 sound card and use its analog line out to a suitable amp/speakers. Not recommended as this may pickup noises inside the PC but may be good enough to test this setup.

Some soundcards with Optical out (which is always 2 channel) come with real-time encoding of 5.1 content using Dolby Live or DTS connect. This can also be used to output up to 6 channels if you have a multi-channel dolby/dts decoding capable amp/processor downstream (like an old AVR)

Best for sound quality is either USB out or HDMI out but what these ports can do will depend on what is connected to them, not intrinsic to these ports. The devices they are connected should provide the capabilities to the PC as part of the handshake on connection so they appear on your list of Sound devices for playback as capable of multi-channel. HDMI connecetd to most TVs, for example, will appear as a 2 channel device only. The same HDMI going through an AVR or a multi-channel DAC will come up with more channels when connected.

So first, pick the output you would like to use. Check its properties as a Playback item to ensure it lists multiple channels in the General tab not just L and R. Then click on the Configure for that item to get the Speaker configuration panel for the device. You should be able to select the actual speaker set up you have (5.1, 7.1 etc). Connect the rest of the downstream audio chain to your speakers and test by clicking on these speaker icons in this configuration panel to get the sounds output. Until you are set with this part, it doesn't make sense to go the next steps.

Select full/large speakers in Windows speaker configuration above so that Windows does not try to create crossovers.

Remember the name of this device in the list as you will be using it later in the configuration of VoiceMeeter. You can even rename it in its properties to be specific. Sometimes, when you have multiple sound devices some of them will have very similar sounding names and you may land up wasting a lot of time picking the one that isn't connected and wondering why no sound is coming out!

 

[Vasr]

2. Setting up Voicemeeter Banana as the core sound engine
You can download and install Voicemeeter Banana from
https://www.vb-audio.com/Voicemeeter/banana.htm
It is donationware. So, it is free to download and use but please do consider providing a donation as it is an excellent piece of software and involves a lot of work.

The user guide for it is at
https://www.vb-audio.com/Voicemeeter/VoicemeeterBanana_UserManual.pdf

It is modeled as a software mixer console for Windows so it looks daunting and the terminology is the same as used for mixers (buses, patches, monitors, etc). I will just list the minimum steps you have to do. You can ignore the rest or wait until you have the basic set up working to tweak it further.

You will need to reboot after installation.

1. Bring up Voicemeeter and click on Menu. In the menu check Auto-restart audio engine and check Windows Tray (Run at startup). This will ensure that it will always startup out of sight on a reboot.

Tip: After you are done with the interface you can use the close windows (not minimize) to get it out of sight. It will remain on the Windows Tray when running and you can bring it up again from there. To shut it down you can use the Shutdown command in its menu.

Note the System settings and option item further down in this menu. You will need it later. For now, let it all be at default values. Also note the Restart Audio engine command at the top. You might need to do this sometimes when things change in the audio devices or HDMI device you are using as an audio device goes to sleep and comes back on. Voicemeeter has no way of knowing when HDMI handshake is re-established.

Tip: If at any time during set up or use, no sound is coming out, restarting the audio engine from Voicemeeter menu is always a good thing to do before debugging further.

2. Click on Output A1 at the top right to pop down a menu of physical audio devices on your PC. A1 to A3 are the equivalent of Mixer outs and can be attached to physical devices. Just do A1 for now to pick the actual device you are going to use for output in Step 1 above. You will find that each device is listed separately for each type of driver available for it. Pick MME to the device you are using. It is not the cleanest and it goes through the Windows Sound Engine, but it is the most foolproof for now for set up and you can tweak it later to use WDM or ASIO if available. Now your "mixer" is set up to send the mixed audio out (although you will be doing no mixing in this entire set up).

3. Voicemeeter has two virtual inputs - One just named input and another named aux input. This is where your audio will be coming into the mixer from external sources. Ignore Aux for now, you will be using the Input channel only. Locate the column for this in the Voicemeeter Banana console. Leave the channel fader slider for this at 0db and make sure mute is not turned on (i.e., it is not red) for this channel. The audio coming into this can be routed to any or all of the Outputs A1-A3 by checking one or more of the vertical list items next to its fade slider. Check A1 to route the input to the Output that you have connected to the external device. Do not disturb any of the other default settings.

4. Go back to the Windows Sound devices list. You will now see two new virtual devices there from Voicemeeter. A Voicemeeter input and a Voicemeeter Aux input (corresponding to the two virtual inputs above). Ignore the Aux input for this set up. Tap the Configure for the Voicemeeter input device item and go through the same procedure as you did for the physical device - set channel configuration and speaker set up. Do the same speaker sound test here as you did earlier but for Voicemeeter. If you have set up the Output A1 correctly in the previous step and directed the Voicemeeter input to A1, sound should come out just as if you had tested on the configuration of the physical device earlier. If not, you cannot proceed to the next steps.

When you do the above test, you will see the volume meters light up in two places in the Voicemeeter Banana console window. One above the channel fader slider for the virtual input next to the square box. This shows signal coming into the virtual input. The second is in the Master section next to the fader slider for output A1 (which should also be left at 0db). There are mix options above the slider. Make sure it is in Normal mode and do not touch it after that. You can click on it to cycle through all possible options but you don't need it.

Tip: Double-clicking on the buttons on the sliders will reset them to 0db if you had moved them.

These volume indicators will be very useful to debug the entire process so you can figure out if the problem is with input or output. There is one vertical bar for each of the channels up to 8. Each of them should light up for each speaker that you test.

If you pass this test, and each channel is sounding correctly in the corresponding speaker for the Voicemeeter input, you can proceed to the next step of installing and configuring.

5. In the Windows Sound playback devices list, select the Voicemeeter input device as the default device. So all programs that output to a default Windows sound device will now go through this. You can now test the set up for now by playing any media player or Windows sound and they should all go through Voicemeeter through to your downstream audio device to come out of the speakers. So far, you have just set up the pipe for this to happen but aren't doing any processing in the middle. So, it will be the same as playing straight through to the output playback device.

Tip: In some cases, you may find that when you pick a 5.1 surround speaker configuration, the side surrounds don't produce sound in the speaker configuration test above and remain silent but if you pick 7.1, then they do. If this happens, just note it down and this remapping of side and rear surrounds can be done later in Equalizer APO.

 

[Vasr]

3. Setting up Equalizer APO to work with Voicemeeter
You can download and install EAPO from
https://sourceforge.net/projects/equalizerapo/

This is again an excellent piece of software that is very stable and works well with very little latency to install eq filters, set up channel level and delays and to create crossovers. It also has features to do some simple channel remapping as well as mixing.

This is one of the Windows programs that ask you to reboot if you sneeze at it for any change so be prepared to reboot at many points.

After installation you should find two specific programs in your Windows menu under Equalizer APO program group.

The first one is the Equalizer APO configurator that you use to specify which devices (physical or virtual) the Equalizer APO should attach itself to. You will rarely use this after the first use. It needs admin privileges to set up.

1. Start the EAPO Configurator and check select the item for Voicemeeter Output A1 which should be listed as the default device (unless the installation itself did this for you) there. You may be asked to reboot after finishing this step. This step makes the EAPO insert itself to the output of the Voicemeeter Virtual input for all audio coming out and before it is sent to the physical device. The processing in EAPO done later in this guide will be applied before sending it out to the selected physical device. In the default fresh state EAPO does nothing so you should still be able to play any sound as in the last step in Voicemeeter and have it coming out of the speaker with no difference. Just make sure you don't hear any obvious distortion as that may indicate some sample rate mismatches in the set up so far (there is a bug in either EAPO or Voicemeeter where sometimes this happens especially when you select a direct mode driver in Voicemeeter to the output device but with the selection of the MME mode as suggested earlier this should not happen).

Tip: If this is set up correctly and possibly after a reboot or restart of Voicemeeter, you will see the letter A inside a rounded square next to all the Rs at the top left of the Voicemeeter Banana console title bar. This will be the only indication that EAPO and Voicemeeter Banana are set up to work with each other.

2. The second EAPO program is the Configuration Editor that you will be using a lot during this set up. So I would recommend pinning it to your taskbar for quick access. When you start it, it will come up with a do-nothing configuration with a pre-amplification gain control (which is like a master volume for the entire thing). Leave this disabled or at 0db. If you had not run Configurator before you start this one, it will offer to do so so that you can select which device it should attach itself to. This step is necessary.

Tip: Sometimes, after a device driver update or other changes in the devices, the EAPO Editor when started will come up with a message that a registry value needs to be updated. Allow it to do so and reboot. So far, I have not seen any adverse effect of not doing so immediately.

3. In the top left for Device: in the EAPO Configuration Editor pick Output A1. It will be in the pop-down menu of all items that have been selected for EAPO to attach to. The channel configuration for this device should be automatically picked up as 7.1 (from device) regardless of how many speakers you have chosen in the speaker configuration.

Some key things to understand for the use of EAPO
a. The processing is applied sequentially from top to bottom of this editor page in that order. Remember this. For all processing applying to the same channel, the order matters.
b. It starts with a basic do-nothing config file. The config.txt file will be in the Config subdirectory of EAPO installation directory. Familiarize yourself with how to get there since you will likely do this often. You can also place a config.txt file in your own folder under your documents and ask it to use that config.txt instead rather than on the System installation area that requires admin privileges.
c. EAPO has a concept of "include" files so that related configurations can be placed into a separate file and then "included" in the main config console. They can also be nested inside them. I strongly recommend doing this for the various different setups to make the management easier. I do separate files for Speaker setup (delay and volume balance), Bass management, REW filters (which have nested include files for each channel as generated by REW). Optionally, you can also have a separate one for any channel remapping or channel copying you might need. So, my main config has just 4 includes and a gain control (as a master volume control for all channels. This makes it possible to switch off entire type of processing (e.g., bass management) by just disabling that include in the main config.

You can add "processing actions" into a config file by clicking on + of the preceding item and selecting the type of action.

We can now get to each of the EAPO based processing of the audio. This is your DSP. Sort of like what you would use a minidsp box for outside.

You will need to install and set up Room Equalization Wizard with a calibrated microphone to do some of the next steps (speaker balance/delay) and for doing room eq with REW generated filters. You can omit some of these steps for experimentation but it will not sound as good if the speakers are not balanced properly in phase and volume especially for multi-channel audio where speakers can be separated by a distance. The instructions for setting up and using REW are beyond the scope of this guide.

 

[Vasr]

4. Setting up speaker balance and delays

This is an optional step if you have a downstream device you are sending the output to (like an old AVR) which comes with its own speaker "distance" or delay and a channel balance setup for each channel. However, these settings are typically not very granular and it is possible to do a more pecise balancing after you set them approximately. If you are not using such a device downstream (straight to amp or powered speakers or DAC+amp) then you will need to do this in software inside EAPO. Without this, any room EQ will not be good especially in multi-channel playing.

Tip: You want to avoid boosting gain for any channel within the digital processing inside EAPO to avoid digital clipping. It is much better to do rough balancing of gain outside this processing downstream so that you only need to do very minor adjustments here and preferably cut the loudest speakers down than boost low levels on some speakers. So, after you do whatever is possible downstream with per channel volume balancing, do the volume balance fine tuning here to bring it to the lowest sounding speaker by negative gains on other speakers. You can control the overall master pre-amplifier gain later in the EAPO Configuration for all channels to bring it to maximum possible without danger of clipping.

Tip: Unlike stereo processing or multi-channel music, multi-channel HT playing has an additional complexity we need to handle. The encoding standards for multi-channel audio in HT specifies the LFE channel to be encoded 10db below that of the other channels. This allows the LFE rumble to go above the reference level in sound production without causing digital clipping in its processing. The assumption is that the volume compensation for this is done at the last stage preferably in analog. Most AVRs automatically do this when they decode and send it to their power amps. But we will not be sending any encoded formats to downstream AVRs if we were using them. Only multi-channel PCM, so the downstream unit has no way of knowing if that 10db compensation needs to be done. If we are not sending it through an AVR, then you would need to handle this in the PC. Otherwise, bass will sound very weak in HT use. But if you were to do this in the digital processing it might digitally clip when it goes above reference level.

On the other hand, you do not need to do this if you have multi-channel music (say 5.1 flac) that is not using that -10db standard for encoding and bass will be booming if you apply this correction.

Note: The above is an issue only if you have multi-channel music with LFE content in it (that is not attenuated by 10db). Stereo music with two channels can be played in this setup with no problems and will play with the correct volumes in the mains and into the sub if you set up the crossover to go to the sub.

There is a way to cater to both multi-channel music and HT content by piping all HT content through the input channel and all music through the input aux channel and setting up the latter to not do the 10db boost (or alternatively do 10db cut on the LFE channel if there is a downstream 10db gain for example in a sub volume control). I will leave this configuration out for now. It will become easier to extend it once this process is understood for HT content with the 10db lower LFE coming through.

The solution I use is to set the powered sub volume 10db higher than other speakers for the same input level to all speakers when I do volume balancing. That way I do not need to provide any digital boost inside the digital processing. The channels going out of the PC will have LFE at -10db and will be boosted back into balanced level by the sub volume control. This has some implications in the crossover design that I will mention. For music, the bass will sound too loud if using a 2.1 or higher and that can be handled by a different path through Voicemeeter.

1. Create a blank file config_speaker.txt (or whatever you want to name it) in the config subdirectory (where config.txt is). All of these are simple text files that can be edited with Notepad if necessary.

2. Add an item into the main EAPO config.txt editor tab by selecting + at the end and choosing Control->Include. This will insert a line of action in the config.txt main tab for including a configuration file. Click the folder icon and choose the blank text file you created above. If you click on the arrow next to it, it will open a new tab in the editor for the included configuration which will be empty at this point.

3. In EAPO, you can specify an action to apply to a specific channel or a selection of channels or all channels. By default, it will apply an included action to all channels. You can narrow it down by adding a channel selector action before specifying the action. The actions below it will apply to the selections in that selector unless it hits another channel selector in processing in sequence from top to bottom.

So, at this point create a section for each channel by adding three things per channel
a. A channel selector Click + -> Control -> Channel selector (select just one channel in each one by clicking the appropriate button)
b. A Preamplification Click + -> Basic Filters -> Preamp
c. A delay Click + -> Basic FIlters -> Delay

Tip: Be extra careful to ensure that each one of the sections above uniquely selects a single channel. If none are selected (or that selector item is disabled) the actions below it will apply to ALL channels. This can create unexpected results that is difficult to debug.

Leave both delay and preamp at 0 AND disabled (by clicking the power button for that item to toggle enable/disable) to start. Do this for all channels. You can turn each one back on as needed when you do the balancing.

4. Set up REW to take volume and reference measurements (for delay). You will need to have installed ASIO4ALL to do multi-channel set up with Voicemeeter channels.

In the REW settings->Soundcard, pick asio as driver, in the ASIO configuration make sure the VB-Audio Voicemeeter VAIO -> Out: .... is checked. This is a critical step that is most easily missed.

You can now select Voicemeeter vaio 1-8 in the output to measure each channel individually. Test each channel in REW to make sure you can get sounds from each of the channels to which you have attached speakers to. Note down the channel numbers here and which speakers they correspond to.

By default, the assignments should be Vaio 1 - L, Vaio 2 - R, Vaio 3 - C, Vaio 4 - Sub, Vaio 5 - SL, Vaio 6 - SR, Vaio 7 - RL Vaio 8 - RR but assignments can vary depending on a number of things in your set up. Ignore the channels that you are not using.

Attach the measurement mic and select that as input.

5. Open the Signal generator and SPL meter tools in REW.

In the Signal generator select Pink Noise PN at the default -20db (you can increase this if you need higher volume to measure). For the sub you will pick Sub cal and for other speakers Speaker Cal (this is where you will need to know the number for the Sub). Select a speaker to send the pick noise to as vaio 1-8 and start the noise generator.

Place the measurement mic at the Main Listening Position (MLP) roughly where your head would at ear level.

Turn on the SPL meter tool to take the volume measurement. You are doing relative measurements here so the absolute number is not that important. But it should be loud enough to be 25-30db above background noise floor.

Measure each attached speaker channel (leaving out sub for now) and note down the levels. Pick the number that is the lowest (least loud). You will cut others down to that level using the preamp setting for that channel EAPO editor panel you set up earlier.

For each channel you need to cut down the volume of, enable the preamp action under that channel selector (be careful here to select the correct channel section to change). Add negative gain (you can use the knob icon anti-clockwise or the down arrow in the db indicator) while playing noise in REW for that channel. Reduce until it is matched with the lowest speaker volume you recorded earlier. Do it for each of the speakers so that they are all at the same volume level.

Prepare the sub by putting the volume on the sub at about 10 o'clock position, and crossover selection in the sub to allow max bandwidth to come in (usually around 120-140hz max).

Now select the Sub for output of the Pink Noise with Sub Cal (important!) and start the noise through the sub. Make sure the input level in the signal generator is the same for all channels including the sub.

Measure in the SPL meter. Adjust the physical volume on the sub until the SPL meter in REW reads 10db higher than the value used for other speakers. This may turn out to be too loud while measuring for your set up. If so, you can lower the volume of other speakers with a master volume for all or lower the input level of the noise generator and do the settings for the lower setting across all speakers with +10db for the sub.

This completes the channel balancing for each of the speakers to get their levels set.

6. For the delay matching, note that you can only add delay but not speed up any channel! So you need to match to the farthest speaker to add delays to the others.

The process is the same as a frequency sweep measurement in REW (Measure icon) using an Acoustic reference (select in the measure window at the top) with one of the speakers.

If you have a device downstream of the PC like an AVR, and it has delay configuration (typically in distance), set it approximately to eyeballed or measured distance from the MLP).

Pick the one that you think is the farthest as reference and measure each of the others except the sub with a sweep. When it finishes and you get the sweep window, the listing on the left hand column will specify a delay and distance measure for that channel as a delta from the reference. Negative delay means you must add that much delay to the channel to align with the reference. Do this for all channels, and list out the reported delay numbers. If any of them are positive, pick the highest number and use that as the reference speaker to do measurements against with all the others. Now you will see negative delay numbers for all of the measured channels. Note them down.

To align all of these speakers, you go back to the EAPO editor in the speaker settings tab and for each channel that has a negative number, enable the delay control and enter that delay as a positive number. This should be good enough for most purposes and much more accurate than setting distance in feet.

If you have OCD, you can go and measure them again and keep iterating (because of measurement errors and variations) until you get the measured delays down to as low as you can. I am able to get the reference level delays in reported relative acoustic distance down to 10mm or less but that is just being obsessive. Won't make much of an audible difference being off a few centimeters or inches.

After a crossover is introduced you can do another round to align the sub and speaker phases but that is out of the scope of this guide.

Calibrating the delay for the sub is problematic because the reference level is at a higher frequency than the range or sweet spot of most subs. So, it is approximate at best. You can try increasing the measurement at the loudest you can bear for the sub while keeping its crossover at the maximum. If the sub allows some mids also to come through, you may be able to get REW to measure the delay delta. But mostly you might find the measurement is wildly off. If it seems reasonable use it. Otherwise, don't bother.

This completes the alignment of speakers for distance and volume across all channels.

 

[Vasr]

5. Setting up bass management

If you are not using a sub OR have large (full-range) speakers and so do not require any crossovers, you can skip this section.

I am not an expert on crossovers and it does seem to be prone to philosophical debates on what is best. I will just note the general process for a simple Butterworth 12db/octave crossover. You can then vary it for any of the standard crossovers you like.

Adding a crossover is very simple. But to really understand this, you must distinguish between LFE and a sub. The nomenclatures in EAPO or in a lot of write-ups conflate the two and in doing so confuse a lot of discussion on crossovers and what to send where. I find it easier to think of LFE as the low frequency channel of content and sub as the physical speaker that will reproduce the lower frequencies in your set up. In an 7.1 or 5.1 channel set up you are simply specifying what should go to each speaker including the sub as the output from the PC. The crossover portion decides what content should go into each speaker given the incoming channels including LFE.

It is much easier to understand the process with the above in mind.

Specifying a crossover in EAPO consists of
1. Copying the LFE channel and the content of other channels into the sub channel for the output
2. Placing a low-pass filter on the sub channel to remove the higher frequencies from the non-LFE channels copied into the sub-channel
3. Placing a high-pass filter on all the non-sub channels to remove the low frequency content.

Since both input and output are referred to as LFE in EAPO it can be a little confusing but keep the above distinction in mind.

There are a few twists to the above:

1. Note that we were expecting -10db attenuation in the incoming LFE channel for HT content and compensating by raising the gain in the sub to play 10db higher. So the LFE content coming in can go straight through to the sub channel. But for the rest of the channels which are copied into the Sub channel, they will be 10db louder and shouldn't be subjected to the 10db gain downstream in the sub. So, the way to handle this is to attenuate the content from non-LFE channels copied into the sub-channel by 10db. This is very easily done in EAPO as shown below.

2. The crossover frequency is typically set at around 60hz-100hz for most common speaker combinations. It depends a lot on the speakers and sub. You will need to experiment or just use the most common 80hz. However, while LFE channel content rarely has any content above 100hz, it may have some content over 80hz, so a steep 80 hz crossover will remove some of the content that should have gone to the sub from the LFE channel. You can avoid this by using a steep 100hz crossover for both the high-pass and low-pass filters or also copying the LFE into each of the other channels before applying the high pass to them.

My suggestion is to try any of the following and see which sounds better or measures better when you do Room Eq later on the crossed over speakers.
a. 2nd order Butterworth 12db/Octave at 80hz crossover. Single high-pass and low-pass filters with a Q of 0.7071 (or using the option for fixed Q in EAPO filter spec)
b. 2nd order Linkwitz-Riley at 80hz crossover. Two identical high-pass and low-pass filters each in series with a Q of .5
c. 4th order Linkwitz-Riley at 100hz crossover. Two identical high-pass and low-pass filters each in series with a Q of .7071

The caveat for c is that you need a sub that can change phase by 180 degrees as the cascade of those two shifts the phase between high-passed content and low-passed content. But that will put the LFE content out of phase by 180. There are ways to handle that but outside the scope of this guide.

3. To start create a new empty configuration file bass_management.txt in the config folder and include that in the main config.txt tab below the speaker management include in the same way you did the latter earlier. Open a tab for this file by clicking the arrow icon in the include item and add the following actions

a. Copy content to Sub channel
Click + -> Basic Filters -> Copy between channels
This puts a graphic interface to connect the input and output for copying but the easiest for what we need to do is to click on the notepad and pencil icon for that item which allows you to enter the text equivalent of the action and enter the following exactly (The Copy: is part of what you should include also):

Copy: LFE=LFE+-10db*L+-10db*R+-10db*C+-10db*SL+-10db*SR+-10db*RL+-10db*RR

If using less than 7.1 configuration you can leave out the terms for the channels that will not have content (your upstream source should be down mixing into the configuration you have). You are copying the LFE content as is and the rest attenuated by 10db. Because of the head room in the LFE content, this should not create any clipping after copying in practice.

If you click on the notepad/pencil icon again, you will see the graphical representation of this to verify

b. Add a channel selector for LFE
Click + -> Control -> Channel selector and click the LFE button in it.

c. Add a low-pass filter for LFE
Click + -> Parametric Filters -> Low pass filter. Set corner frequency at the crossover frequency and leave it at Fixed Q if you need a Q of 0.7071 or select Q Factor instead of Fixed Q and enter the Q value you want. If doing b or c above do this again as another filter below this with identical values to cascade the two.

d. Add a channel selector for all the other channels copied into LFE above.
Click + -> Control -> Channel selector and click to select all the buttons for the channels copied into LFE above. The same action below this will apply to all the selected channels.

e. Add a high-pass filter for these channels
Click + -> Parametric Filters -> High pass filter. Set corner frequency at the crossover frequency and leave it at Fixed Q if you need a Q of 0.7071 or select Q Factor instead of Fixed Q and enter the Q value you want. If doing b or c above do this again as another filter below this with identical values to cascade the two.

If you are doing dual subs as 5.2, you can use the above guide to extend the copying and filters to the relevant channels you are using for the two subs.

Your bass management with crossovers is done!

If you are OCD, you can do a REW sweep again with the crossover in place and check the phase results and fine tune the delays for the speakers to align phases. You can also tweak the crossover frequencies of the low and high pass filters and/or tweak the slopes if you find issues in your crossover region like nulls, etc. How to do this is beyond the scope of this guide.

Go ahead and play some content (stereo music or HT content) and see if the sub integration is working well for you (albeit without room eq as yet).

 

[Vasr]

6. Adding REW generated room-eq filters

The procedure for creating the EQ filters for each speaker using REW measurements is beyond the scope of this guide. For the purposes of this set up, the EQ filters are to be exported from REW into individual files for each speaker.

While there are several philosophies on how to generate room eq when a sub and crossover is involved, the simplest and most practical approach is to measure each speaker independently with the crossover applied and generate filters over the entire range you want to EQ including sections that will be crossed over. Since the Room Eq filters will be set up in EAPO to be applied for each speaker before being crossed over but measured with crossover applied, the corrected curve will be fine when the crossover is done after applying the room eq.

Once you know how to generate the REW eq files (as Generic Equalizer export to text file in REW), it is trivial to include them in EAPO.

Tip: When you measure each speaker in REW, you can name that measurement in the left hand column that lists each measurement. Instead of the time stamp typically used by default, name each measurement you will use to EQ in the format

Channel: X

where X is one L, R. C, LFE, SL, SR, RR, RL depending on the speaker used. When you do this, the export to text file for the filters will contain the above line at the beginning which in EAPO will select that channel to apply the following filters. So the exported file can be used as generated by REW with no modifications! Nothing could be simpler.

So, the steps are:
1. Create a blank text file in the config subdirectory named something like rew.txt. Include this file in the main config.txt tab between the speaker level/delay setting include in Step 4 above and the bass-management setting include in Step 5 above. To do this click on + on the speaker level/delay include item and select the blank rew.txt just created. So, the order of includes so far as they are processed from top to bottom are - Level/Delay settings, REW filters, Bass Management.

Tip: You can change the order of items in the EAPO editor by clicking and dragging the tab at the left edge of the item up or down.

2. Click on the arrow button in the rew.txt include item to open the blank rew configuration in a new tab in EAPO editor.

3. Generate the individual EQ filter files for each channel using REW and save/copy them to the config subfolder.

Tip: Save these files with the name of the channel and a date stamp so that it is easy to identify, modify and update individual channel files later

4. In the EAPO editor tab for the rew filters (rew.txt), add an include item for each channel
Click + -> Control-> Include and select a REW generated channel filters file to include. If you followed the first tip above, these files will already have the channel selector in them so EAPO knows which channel to apply the filters from each file. You can verify this by opening each of the included files in a tab. Make sure this channel selection happens, otherwise, it will apply the included filter to all channels.

The order of the above includes for channels does not matter.

Now you have REW based EQ enabled in your audio chain!

I recommend two more fine-tuning steps after this.

1. Go through the volume setting procedure above in 4 again with the eq and bass management enabled and level the volumes of all the speakers again using the directions in step 4 above.

2. Remove any potential for digital clipping.

To do this in EAPO, you will use the Analysis area at the bottom of the editor window. The left side allows you to select each channel in turn and see the corrected curve (for a flat input) on the right based on all the processing done in your set up. You will see the effect of the bass-management filter as well.

Make sure the left side says start in config.text, your starting configuration file. Select each used channel in turn in the menu on the left hand side and see if any of the output levels go above zero (will be displayed in red below the selection). The right side will also mark the areas of the curve for that channel going above 0db in red.

If you have any channel showing output greater than zero, note the one showing the highest positive value in red.

Go to the main config.txt tab and enable the master pre-amplifier gain item that came with the original blank config.txt. Turn the gain down by the amount noted above. You will see the output gain go down to zero for the channel that showed the maximum value over 0db. All others will now be at or below 0db and you will not have the possibility of digital clipping.

Tip: You can enable/disable the RoomEQ by just enabling/disabling the include (click on the power button for that item to toggle) for REW filters created in this section.

Tip: EAPO detects changes in the configurations in real time and applies them automatically if the Instant Action box is checked on the top left hand side of the EAPO editor. Caution: If you have this on and you include a configuration that happens to have a very high gain applied for one or more channels and playing something at the time, you can blow your speakers. This is unlikely to happen if you are careful. If you don't trust yourself, turn the Instant Action off, make any changes. Look at the result curve for each channel and ensure that none of them have any high gain in red for output before you check Instant Action to apply them.

Tip: If you want to change the eq for any speaker for a new measurement, you can just copy over the old file in the config subdirectory for that channel while EAPO is in the Instant Action mode and it will automatically use the new EQ. This makes it easy to do quick A/B testing. I do recommend some version control for these REW files since sometimes you will want to go back to a previous setting.

Go ahead and enjoy your stereo music or HT content with this chain by pointing the player to output to default Sound device (set as Voicemeeter input VAIO) or point them to the latter explicitly. Your movies will sound much, much better than without all of this work. The music will sound clearer and crisper without any room induced boominess or muddiness.

If you have done the REW filters well, the center channel dialog will be clean and crisp, the transitions between your speakers for the sound will be much smoother for an immersive experience and you will get the maximum impact of the low frequencies as much as your subs will allow without any room induced boominess.

 

[Vasr]

7. Adding multi-channel Dirac Live 3.x as an alternative

Adding Dirac Live Processing to this chain instead of using REW filter is fairly easy (after one has figured out how!). It uses a not so well documented feature of Voicemeeter and some gotchas in using Dirac which has only recently graduated into handling multi-channel content (with the HT quirks) from its music roots. This is as a stand-alone processor on a PC using DLP, where they don't have hardware vendors implementing that part incorporating Dirac Live in their AVRs.

In terms of understanding the Dirac architecture, it consists of two parts. The Dirac Live software that is like REW that is simpler to use for measuring and creating filters and the Dirac Live Processor which is the "soft DSP" that applies the filters to a stream of music and does what the EAPO embedded REW filters did above.

The complication is that DLP comes as a VST plugin (https://en.wikipedia.org/wiki/Virtual_Studio_Technology) and it is not obvious how to use it if you are not familiar with VSTs. If you are using certain commercial software that can use VST plugins, then it is simple as they will manage its use. But in this generic architecture, it requires what is known as a VST host that knows how to handle VST plugins and a mechanism to insert it into the audio chain. The VST host associates a playback interface to the VST input interface and a listen/record interface to the VST output interface. So, externally the media player plays to the input audio device and the processed stream comes out of the recording audio device.

For two-channel music, DIrac recommends a Virtual Audio Cable from the same developer that created Voicemeeter. This virtual cable creates a playback audio device and a record audio device that are connected. Which you use as the input and outputs to the VSTHost. But this does not work for more than 2 channels content on Windows (after XP). This is because in current versions of Windows, a recording device can only have 2 channels unlike a playback channel that can have eight.

But there is a way to use a VSTHost as a "patch" in the audio chain using Voicemeeter and that is the solution used here.

Issues:
1. DLP, as of current released version, does not have bass management yet, so it cannot do crossovers. So, for this alternate solution, we will keep the EAPO in the chain and disable the REW processing (by disabling the include file for the filters if you did the previous step) but keep the bass management part. So, to Dirac it will look like up to 7 full range speakers and a Sub for the LFE.

2. DLP, as of current version, also does not understand the 10db attenuation of HT content streams and so Dirac Live measurements don't work right if you have configured the sub to be 10db louder in the set up above to handle HT LFE correctly. Its testing will use the same level for all speakers including the sub. The sub will sound louder and therefore Dirac will attenuate the sub in its configuration for the DLP. But when the already attenuated LFE stream comes through this setup, then it is further attenuated and so bass will sound very weak. So weak that the HT experience will be really bad.

On the other hand, setting up the chain to remove the LFE attenuation from HT content prior to Dirac will do this in the digital domain that can lead to digital clipping as it may go over 0db. My work around is to remove the +10db gain in the sub before doing Dirac Live measurements and after the filters are in place in the DLP. increase the sub gain at the sub by 10db as before.

3. Dirac will do speaker level balancing but it works better if the speakers are already in (or close to) balance so I would leave the speaker level settings in EAPO as-is.

4. Do note that Dirac will reserve 12.5db of headroom for its processing so all your streams will play at lower sound levels with Dirac processing on. It can be compensated if you have a volume/gain control downstream for all channels or wherever the digital to analog conversion is done.

The steps to do the Dirac incorporation are:

1. Disable the REW filter processing in EAPO if you did the earlier step. You can just disable the include line for that configuration (rew.txt) in the step introduced for incorporating REW in the previous stage.

2. Turn the sub gain down by 10db if you had configured it higher in Stage 4 above.

3. Download and install Dirac Live 3.x and DLP both (as a limited time trial or a licensed version). Note down the directory DLP installs its plugins during installation.

4. Download and install a VST Host. There are two free ones you can use (there may be more)
VSTHost (https://www.hermannseib.com/english/vsthost.htm) or
Cantabile Lite (https://www.cantabilesoftware.com/free-vst-host)

VSTHost is an ancient, bare-minimum interface but much easier to set up and going than Cantabile Lite but the process is fairly similar. I will use VSTHost here as the example.

5. Run VSTHost after install.
In the File menu, check Plugin Auto-Start. And then select Set Plugin Path in the File Menu. In the dialog box, you will add the path to where DLP installation installed its plugins.

Tip: DLP comes with plugins for both VST2 and VST3 versions of VST technology and are maintained in separate folders. Dirac VST3 plugins did not work for multi-channel stream, was able to only output stereo. So pick the folder location for the VST2 plugins above.

VSTHost will scan the folder and the list of plugins will now be available in the Plugins submenu under File Menu (not Plugins menu). It will list Dirac Live Processor there. Select DLP. You will now see a small window for DLP appear in the VSTHost workspace and its existing engine input and engine output will be connected with DLP in the middle. This is automatic and you do not need to do anything to make this connection happen. DLP will verify your license with Dirac servers here, so you have to be connected to the Internet for this to work whenever you start VSTHost and your license needs to be valid.

Tip: You can use the Bypass command (or Alt+B) in Plugin menu of VSTHost to bypass the DLP plugin or insert it back. This is useful for debugging to ensure that the VSTHost is working in the chain without using the DLP plugin.

6. Insert VSTHost into the audio chain set up earlier using Voicemeeter.
This is a slightly complicated process with multiple steps. The principle is that Voicemeeter (like a mixer) allows a "Patch" to be inserted for each input channel (hardware or virtual inputs - each of the mixer channels with the slider inputs to the left of the Master faders). The patch is an external processor to which the incoming signal is fed and the return from the processor is then used as the input to the mixer. We will configure VST Host as the external processor patched into the Voicemeeter input virtual channel we have been using for input.

a. Click on Menu in Voicemeeter and select System Options/Settings. Ignore most of what comes up except for the very bottom which has sections for each input with what looks like patch cables for each channel. The hardware inputs are all stereo (Windows limitation) and so two patches. The virtual inputs have 8 patches each. They are in the same order as the channels in the mixer interface. So the patch section for Input virtual we are using is the middle 8 patch section labeled In #4 Virtual input. Click on each of the 8 boxes in the middle to turn them on. The two squares and connecting lines turn bright green showing the patch is enabled for that channel.

b. To connect the VST Host as a patch we use a virtual audio cable that the installation of Voicemeeter uses with an ASIO interface (so that the two channel limitation of Windows is eliminated). This cable device is called Voicemeeter Insert Virtual ASIO but it does not appear in the Windows sound devices listing.

c. In VSTHost, select Wave in Devices menu.
In the dialog box, skip the input device (this is different from other uses because of ASIO) and for output select the ASIO: Voicemeeter Insert Virtual ASIO from the list. Leave the sample rate at default for now. Close this dialog

d. Select Configure from Engine menu of VSTHost.
Here we have to do the mapping of the VSTHost engine inputs and outputs to the Voicemeeter patch cable channels. You will do the same thing in both Assign input and Assign output tabs. Select the number of channels you want to map (say all 8). It will create that many entries below. For each entry click on ... button to select the Voicemeeter channel. You will be using IN#4 channels for the virtual channel we are using with the same labels as on the patches in the Voicemeeter. Use the same order of channels and it should work. If you find that Dirac Live is sending a measure signal to the wrong speaker while measuring you can come here and reassign. Make sure you use the same order for both input and output assigns. This should be L, R, C, LFE, SL, SR, RL, RR in most cases.

e. Click the Plugin editor for DLP (Dial icon in the top right of the DLP window inside VSTHost). This shows the current filters installed. It should be empty for now. In the top right corner of this editor choose the speaker configuration you will be using Dirac Live to measure (5.1, 7.1, etc). Use the configuration that corresponds to the actual speakers you have as Dirac Live will want to measure each one of them.

VST Host is now ready for use with the (empty) DLP processor patched into Voicemeeter.

Test by playing some content like before to the default/Voicemeeter Input device selection. The sound should come out as before with DLP doing nothing. You should see the volume meters light up in the input engine, output engine and DLP windows inside VSTHost for the number of channels that have sound.

6. You can now run Dirac Live to do the measurements and it will do the installation of the filters into DLP. Dirac suggests that you keep the path playing with content before you do Dirac Live. Dirac Live will mute the sound being played when doing measurements.

When you start Dirac Live, it will detect DLP as the device that is Dirac Enabled and do the measurements for that device using the number of speakers you specified in (e) above.

The exact process for using Dirac Live measurements is beyond the scope of this guide. But once it installs the filters Dirac processing will be active in the chain for any content you play through. Enjoy!

You can save the program in VST Host to save the configuration and add a shortcut to the VSTHost program in your Startup folder so it automatically comes up on a reboot.

 

[Vasr]

8. Adding an external source to the PC via optical input

You can add an external device with Optical output (CD Player, TV Optical out, etc) into the set up above to send the sound to your speakers with bass management, room eq, etc. The limitation is that it only works with 2.0 channel content so that it comes via optical as unencoded PCM (no Dolby, DTS, etc).

For each such connection, you need a PCIe slot and install a sound card with an Optical port in. Not many cards exist for this. Soundblaster AE-9 is an expensive solution. There are two white-labeled inexpensive cards I have found that work well.

https://www.amazon.com/StarTech-com-7-1-Channel-Sound-Card/dp/B00X7B2NY0
or
https://www.walmart.com/ip/Syba-7-1-Surround-Sound-PCI-e-Sound-Card-S-PDIF-In-and-Out/35742487

They are the same cards rebranded. You don't use it for anything else than the optical in and it works out of the box without any driver installs. The driver card isn't necessary unless you want to use its DAC for analog out. So you can add another main card if you need two ports.

1. Install the sound card in an available PCIe slot on the PC.

2. Windows will install the default drivers and make these available as sound devices in the Sound devices panel.
Under the Recording devices tab you will find the Optical input in device. Remember its name.

To play through the existing setup as set up earlier above:

Double-click on the optical input device for its properties. In the Listen Tab, check Listen to this device and select the Default device (if you set Voicemeeter input as default device) or Voicemeeter input device directly. Apply. This will play the input 2 channel content coming through the Optical in with all the bass management, REW/Dirac processing set up earlier.

Alternate Solution:

In Voicemeeter, click on Select input device under Hardware device 1 and choose the Optical input identified above. Use MME driver again for now.

Turn on A1 to connect it to the output in use. This will send the input from the Optical in to the output to which EAPO is attached to use its processing including REW room eq if it is enabled. However, note that Dirac "patch" is attached to the Virtual input and won't be applied to this flow. This is a limitation of this approach. I have not tried to set up multiple instances of VSTHost to patch the Hardware input as well to see if Dirac can work in this method.

 

[Vasr]

9. Miscellaneous Topics

1. Alternate processing based on content (e.g., music vs HT)

Briefly, it is possible to set up separate processing for different sources. In the set up above we attached EAPO to the output A1. So any sound going out to that output will use the EAPO processing.

It is possible to attach EAPO to the input device instead so you can do different configurations for different inputs. To do this, in the EAPO configurator, rather than enable EAPO for the Playback device, you can attach it to one or more "Capture Devices" (the inputs). So you can attach it to Voicemeeter virtual input and Voicemeeter aux. The former should have the same effect as the previous and by default the same processing will also apply to the sounds coming into the Aux Input device.

However, in the editor you can create separate sections for each of the attached devices by using a Device Selector in Basic Controls. All controls below a device selector will apply to the Device selected above it.

This allows you to set up different sources differently. So for example, you can do the existing processing for both input devices but for the AUX input, you can attenuate the LFE by 10db. Now, if you play your multi-channel music with LFE content by pointing your player (say Foobar) to the Aux input device, the 10db gain you had set up in the sub will play the right level for music in which the LFE channel is not attenuated as in HT.

I have not tried this particular configuration.

2. Channel remapping in EAPO for surrounds

As mentioned early on, it is possible that your channel mapping isn't working correctly and no sound may be coming out of your Side Surround speakers in 5.1 configuration. This can happen if within the chain, 5.1 content is being sent to RL and RR instead of SL and SR. If this happens, you can redirect the RR and RL to SR and SL by copying the streams (if RR and RL are not connected). This is done using the Copy control as described in the Bass Management section. Insert it at top before any processing.

In the Copy control, Click on RR in input in top row and drag to SR in bottom row and similarly from RL to SL. You will see two lines that are copy lines. Now anything that appears in RL and RR will appear in SL and SR and go to the side speakers.

If in a 7.1 configuration, the Side and Rear surrounds are swapped, you can swap them back in the above Copy control by also connecting SR in top to RR in bottom row and SL to RL.

 

[phoenixdogfan]

I am going to post some simple notes of the steps below broken up into multiple posts so they can be individually discussed rather than everything in one long post. So, please bear with me until I am done. I will break this up into the following posts:

1. Prepping the PC and requirements
2. Setting up Voicemeeter Banana as the core sound engine
3. Setting up Equalizer APO to work with Voicemeeter
4. Setting up speaker balance and delays
5. Setting up bass management
6. Adding REW generated room-eq filters
7. Adding multi-channel Dirac Live 3.x as an alternative
8. Adding an external source to the PC via optical input
9. Miscellaneous Topics

Pleased don't ask me for basic Windows concepts for Audio or details of how the above tools work. You can pick them up by googling. I am just going to list what you need to do with these tools to set it up.

My use case is going to be much simpler, I will decode ripped movies and streaming video from Netflix, Amazon Prime, and Hulu.

For that I will use JRiver as my multichannel decoder player for all legacy formats up to 7.1. I have original Dirac Live, and have not gotten around to upgrading it to 2.0 (3.0) multichannel, but I will once I get other business finished.

I will send the video from the streaming services to JRiver to decode. It will be decoded to 5.1 legacy audio and sent to Dirac Live for any room/speaker correction. From there the video will be sent via HDMI to the monitor and the audio will go to the soon to be ordered Octo Research DAC 8 Pro. From there the audio will go to the amplifiers. Front L/R currently a Purifi Eval 1 + LS 50, phantom center, the sub an SVS Sb2000, and the back channels Mirage Omni Sats driven by my old Behringer A500. The remaining two channels of the Octo 8 will decode the Spidf feed coming off my Smyth Realiser A16 which will be sent via balanced cables to my THX 789 for when I listen with headphones.

In the future, I may upgrade the front L/R to either something from Revel or Kef or possibly even Dutch & Dutch, and move the LS 50s to surround duty.

I've used JRiver for ages, and sent the video to the monitor via HDMI while sending the audio to whatever preamp/processor I'm using via USB and never have I had severe latency issues, but then again I use an I7 with 16 gb of memory and a Samsung Evo SS drive. So no problems with latency.