Solution for Those that Want it: Sending all system sounds + music to audiolense convolver

[mccarty350]

1. Install hificable/asiobridge.
2. Install voicemeeter (take your pick as long as it has at least two inputs)
3. Install audiolense convolver.
4. Install your musicplayer of choice (I use jriver some days, foobar 2000 other days).
5. Set jriver's output to be voicemeeter aux 1 using ASIO. You can set it to the minimum hardware required latency and it works fine.
6. Set jriver to resample everything to your desired sample rate.
7. Make sure that all drivers (hificable/asio bridge, voicemeeter asio in, voicemeeter aux in) are set to that same bitrate and sample rate that pertain to output and do the same for inputs in your sound control panel.
8. Set the output of voicemeeter A1 to VB-Audio HIfiCable.
9. Open asiobridge/hificable application.
10. Set asio destination in asio/bridge/hificable to be 'audiolense virtual sound card'.
11. Audiolense convolver will pop up. Set the buffers to 2048, kill the additional buffers option (at least in my setup). Load your convolver and set audiolense convovler's output to be your asio sound card.
12. Enjoy everything on your system getting convolved! There are bugs but it works.


If you have questions or if there is over arching interest in this I'd be happy to help since I just fought through this for literally hours today trying different buffer sizes and sequences of events, etc. with different programs crashing as I changed settings in others.

 

[Ofer]

Thank you. Did you manage to route an outside source like spdif to Audiolense convolver? I have yet to manage this.
Also with the setup presented, where do you control the volume? At AL convolver? at the music app? Does this setup work for video playback from external apps like netflix (it should)?

 

[mccarty350]

Here are some other gotcha's I'm putting in here for general knowledge:
1. If you point output A1-Ax at the audiolense convolver virtual sound card in 64 bit voicemeeter it will crash. O
2. On the 64 bit version of voicemeeter it will only show ASIO output having a capability of 44.1khz. Run the 32 bit version and both work. This has been noted in other forum posts as well.
3. I actually do the following:
a. Point my entire operating system to voicemeeter vaio 1.
b. Manually point jriver at AUX input and I point foobar at VAIO3. I do this just for easy identification/troubleshooting of source.
c. Voicemeeter output points at the convolver.

The use case I'm solving is the fact that this will let you hear games and system sounds while listening to music at the same time and still have convolution applied to ALL OF IT. Another use case (this is my long term goal because I have a pair of Linkwitz LX521.4's driven by a minidsp that I would like to eliminate/sell) is that if you have active speakers you HAVE to have overarching convolution implemented for ALL sound output because the convolution filter would be acting as the xover and if you don't do this you risk blowing a tweeter (or other driver) by sending it an unfiltered signal. All of these gyrations are in preparation for a future initiative to elminate the minidsp and to (hopefully) use XO to implement the crossover.

I am using an external Sherbourn PT-7010A 8 channel preamp/decoder (you can buy these things all day for under $300 and they have a very good SNR, these products also were rebadged Sunfire products that were 4-6k in their day that are now useless because they don't support HDMI but work GREAT as a multichannel preamp) to compensate for the gain lost when you have a convolver in place and to provide some voltage gain above and beyond what a sound card or audio interface provides. My ASUS Xonar U7 Mk II only puts out 1 vrms so having a pre really helps with versatility. It also has an IR remote and can be controlled by a logitech harmony if you find one that is missing an IR remote.

When you say an outside source give me some more specifics. Are you basically generating your audio multichannel in 5.1 or 2 channel from some other device and then taking SPDIF into a PC from that device and wanting to convolve it? If so then you SHOULD be able to make this work by sending that source to one of the inputs of Voicemeeter and then you assign output A1 of voicemeeter to 'Audiolense Virtual Sound Card'. Behind the scenes voicemeeter operates like VBaudiocable (made by the same company) without all the headache of figuring out how to configure the virtual cables. If you give me more details I might be able to give you some hypotheticals on how to get it going. I just picked up convolver this week and have spend god knows 10+ hours fighting settings to get this work so if I can save you time I'm happy to do so.

 

[Ofer]

Hi @mccarty350,

So the idea is simple (so I thought). To run a signal from my TV/CD to a digital preamp (miniDSP HD10 in my case). From the digital preamp to my RME ff400 spdif in. Have the sound convolved by whatever, JRiver/AL convolver and then sent to the RME outputs from there to my amps and speakers. The system is 3way active xover with 1 sub, stereo only.
BTW why didn't just use JRiver convolver for you playback as all your media is on the computer? Did you try JR WMD driver? it is supposed to do what you want. I haven't used JRiver, all of my suggestions/questions are based on digging (relentleslly) on the web.

 

[mccarty350]

So here's the reason why I'm not using jriver. I'll give you a scenario and it will make it evident (because I did try using it)
1. You're playing audio from JRIVER with the audio pointed to your sound card/interface (by the way, nice interface you have there!)
2. You're playing music from jriver and it's going out your sound card.
3. You point your system sounds and default sound to the jriver wdm driver.
4. The second your system makes a sound it shuts the track you were playing down and jriver starts playing the ipcm stream from your operating system.

In short...jriver isn't a mixer, it will play either the IPCM stream OR your music not both at the same time.

Does that make more sense now on why I've pursued such a convoluted path?

What we are trying to do is apparently not that common and since ASIO is an exclusive mode driver and we need to use it it kills the windows mixing for the most part. Maybe RME has proprietary software that circumvents this but my previous audio interface did not. Turned out that for me a sound card that had 8 outputs did what I needed it to cheaper and easier than the interfaces I tried, RME was not one of them however due to $$$.

So in your scenario I'm guessing you should be able to take your RME SPDIF in and point it at one of the inputs to voicemeeter and then set voicemeeter to point at the convolver.

Does that make sense?

Oh and by the way, what worked on my system earlier that I described to you as a solution after a few reboots randomly is now full of crackles and such. Looks like I'm back to tweaking buffers and parameters across a million pieces of software and settings scattered all over.

 

[Ofer]

Thanks for the explanation. I shut windows sound but as you said JRiver is not a mixer and maybe something else will crash the audio. The RME is supposed to be a muti asio client. Or so it says. At the standalone AL convolver it didn't work, the convolver just didn't detect the spdif input I have never used voicemeeter. Could you be so kind and post some screenshots of your configuration? Also how did you configure windows sound with all this?

 

[mccarty350]

Yeah, it's quite a circus. Just to let you know suddenly things started working and kicking out crystal clear 96/24 audio out of the blue. It is crazy how much randomness is a factor in audio tinkering on a PC. So for now my proposed solution stands to a degree.

Yeah, AL convolver seems to only want to receive things using ASIO. I know there is a stand alone mode but I am yet to fully comprehend how it works. Regardless you have options to get stuff from your source to Audiolens convolver. Here's some things I've tried:

1. ASIOBridge/Hificable. It's a pain the to get it work but it will take an input and you can make the destination the Audiolense ASIO driver.
2. You send all your stuff to voicemeeter which will take anything, wasapi, wdm, ks, asio, etc. on it's inputs and then have it spit all of it out directly to audiolense virtual sound card.

There are probably others but I'm still learning. I suggest grabbing voicemeeter and just giving it an eyeball, it's donationware (it's good so after I used it for a few weeks I donated) to see what options there are using it. I found out about Voicemeeter by posting my issues in JRIVER forums and their staff suggested that I try voicemeeter to accomplish what I wanted to do i.e. have a total overall system mixer that would take anything in and spit it all to a stand alone convolver.

Windows sound......that's another mess right? I basically took every single driver and set it to be 96/24 across the board. I'll take some screen shots and post them for you.

 

[Ofer]

My miniDSP that will serve as a preamp up-samples everything to 96/24 so I will do that way too. What about other internal sound sources. YouTube, Netflix, Chrome, Tidal/Spotify. Do you need to configure a virtual input for each on on Voicemeeter? I will definitely give your setup a try. It makes sense.
Thanks.
BTW I see there is Voicemeeter and voicemeeter banana. Which did you use?

 

[mccarty350]

This may come in as multiple messages as I'm able to break between meetings and post things. I'll also post some of the things I've discovered. You may already know them but hell, anyone could read this thread and gain from it.

1. With ASIO buffers less is more in most cases. It seems like if you put buffers in larger than 2048 you get crackles and things fall apart. If you go too far below that you get he same so it feels like buffers should be somewhere around 2048. This goes beyond trying to minimize latency which I am less concerned about, this is more that it just doesn't work right and from what I've read too big of a buffer is as bad as too small of one. Thus the element of randomness.
2. These applications check each other's capabilities behind the scenes. When I say this I mean that if you make a change in buffer in audiolense convolver while jriver is playing then jriver may crash because it somehow communicates with that software and knows what it's buffer is. So basically at minimum pause your music when you are making changes in audiolense convolver, it looks like it restarts itself (VERY VERY QUICKLY) every time you chagne a critical parameter and if jriver is playing it will crash. Same for voicemeeter.
3. This is super unscientific as in doing the same thing twice doesn't produce the same result which makes figuring things out insanely difficult. So don't dismiss things you've tried before sadly.
4. Uninstalling and reinstalling everything sometimes fixes stuff.


Here's a sad example. I pointed voicemeeter's output at audiolense virtual sound card. Suddenly ASIO4all pops up but not audiolense. I don't know why.

I restart voicemeeter several times and this still happens and audiolense convolver is still not receiving audio.

Suddenly on one random attempt asio4all doesn't pop up and audiolense convolver first up like clockwork and everything is great and I did nothing different that I can perceive.

Anyways, here are my settings at each end of the chain - ignore my default devices, I set everything up discretely so whatever is set to default doesn't matter.

Windows:
Note: This is playback.

Note: This is recording

I'm just going to use one of the virtual sound cards voicemeeter throws in as an example. Set it all to allow exclusive mode, etc. and hard set it at whatever sample rate you want the whole system to operate on. 96/24 is my chosen bitrate. I'm sure that 48/24 would be great too, I use 96/24 for no good reason really that I can rationalize.

Jriver below. The key things to take from this is that you need make sure you are kicking out the same sample rate. I'd also dig into the settings and make sure you're using the high quality SOX resampling. The other piece is that you can output two channel from jriver because the convolver itself I THINK should do all the channel mapping so all you have to do is ship a stereo two channel stream to the convolver.

Here is voicemeeter. Lots of dials and madness make it look complicated. Bottom line are the things I've circled. Set output A1 to Audiolense. The three virtual sound card input drivers I've circled. I labelled them with what I use them for. They all get mixed and the slider bars below them control the levels.

Below are the system settings for voice meeter. All of this is autopopulated except for what I circled. Set it to your system desired sample rate.

Here is audiolense convolver when fed by voicemeeter. You shouldn't even have to do anything manual it should be triggered/pop up by playing sound through voicemeeter which pipes it to audiolense which autostarts it I think.

Last image showing that I set the buffer to 2048 (at least for my chosen bitrate). I have no idea if I need the roon fix since I don't use it and extra convolver buffer your mileage may vary. I've heard it both solve and introduce noise in my tinkering.

Hope that helps!

 

[mccarty350]

My miniDSP that will serve as a preamp up-samples everything to 96/24 so I will do that way too. What about other internal sound sources. YouTube, Netflix, Chrome, Tidal/Spotify. Do you need to configure a virtual input for each on on Voicemeeter? I will definitely give your setup a try. It makes sense.
Thanks.
BTW I see there is Voicemeeter and voicemeeter banana. Which did you use?

I've got two of those so I'm very familiar. I'm planning on eliminating them. Sounds like you and I both use the same system target sample rate so my examples will be spot on.

Other internal sound sources will just work by setting the default system playback - all of it - to one of the virtual sound cards in Voicemeeter. You can pick it as your default sound card so that everything gets funneled there and then the exceptions can be made for jriver and anything else you want 'separated'.

Every 'bigger and better' version of voicemeeter adds virtual sound cards. I think voicemeeter has one input, banana 2, potato 3.

 

[Ofer]

Wow. That is a lot to take in. Thanks a lot. I will try it with feeding the mixer my spdif too.
BTW for me the idea is to go from miniDSP based xover and DRC to audiolence+audio interface xover and DRC. Since using the smart TV as a source is so easy and friendly I don't intend to go fully media computer. So the HD4x10 will be "downgraded" to a digital preamp. Every once in a while foobar will play local files. But the truth is with cd quality streaming it is all to easy to cast/play from the TV.

 

[mccarty350]

It was an insanely painful amount of trial and error to even figure out to articulate it. You'd think there would be an easy way to listen to your music and getitng your system sounds and games while still using convolution but there isn't sadly.

Sounds like the same goal I have, we can get so much more resolution and correction using a PC with unlimited horsepower than we can with the super limited amount of changes we can make via PEQ using Minidsp.

I'm using an NVIDIA shield for all things TV going to an Emotiva XMC-1 that runs dirac but my HTPC also feeds into that XMC-1 so I have my choice on how I do things and the sources i.e. PC based music vs streaming sourced from the shield, etc.

If PC's made life easier I'd do it all from a PC - but they don't.

 

[Ofer]

No it does't. I have installed bannana unfourtunatly I haven't found the spdif option in the hardware inputs only analog 1+2. I would keep digging. How did you change the buffer in AL convolver? mine is stuck on 256 and I can't just dial in another number.

 

[mccarty350]

I know it's not incredibly intuitive but in a flurry of activity by pounding the keyboard and clicking around I discovered that if you hover over the buffer and right click it increases the number and by left clicking it reduces the number. I have not found a way to just type in a value.

 

[Ofer]

Not intuitive, you don't say... I feel I am getting closer. It seems that the way to go with ASIO clients is to add strips it the setting menu of voicemeeter banana, see screenshot. Unfortunately it introduced all kinds of crackles and pops. The spdif shows as sound on the 6 channels of the convolver in the TotalMix FX console but it is all crackles and pops. The computer sounds through the HiFi cable also have crackles and pops. Hovering over the buffer size did not impress the convolver it but it did change when I changed the buffer size in the audio interface, the maximum was 1024 though. It seems that it is being routed to 6 software playback channels instead to the 6 physical outputs by the convolver.
This is some steep learning curve....

Ok. got rid of the cracks and pops by increasing buffer size in totalmix setting to 512. The correct strips were 9,10. Now there is sound through all output channels which basically means that the sound doesn't go through the convolver.

 

[mccarty350]

It looks like you are deep in it and are starting to understand the pieces. I'm guessing you'll figure it out if you keep banging away at it. Like I said I sent a tremendous amount of time even getting where I was but two days running it feels like I'm done and stable even after many reboots.

I'm not using any patches or strips in my configuration, I just have the entire output A1 in voicemeeter dumping to audiolense virtual interface. That triggers audiolense convolver to pop up and in the first context screen at the top left corner it should show like 'voicemeeter8' or some permutation of voice meeter. Looks like you are fighting with hificable there - I did that for a while too but it's SOOO picky and sensitive that I just wanted to eliminate it from any involvement in the system.

 

[Ofer]

So did you remove HiFi asio bridge all together?
Did you get to measure, create and implement filters with AL?
If so how did it compare with Dirac?
BTW your processor looks impressive.
Seems like I hit a dead end (again). It looks like VM doesn't send the audio to the convolver. The miniute I introduced ASIO strips all music plays on all 8 channels instead on the 6 designated in the convolver and there is no difference between the outputs i.e. all are playing all the spectrum (20-20khz). This is true for Hi-Fi cable and VA-cable.
Feels like trying to piece together a puzzle where the pieces to match the picture on the box...

 

[mccarty350]

I did, I totally unstalled hifibridge/asio bridge.

Here's what I did to generate filters for my office fully active system:
1. Defined the speaker setup as 2 channel.
2. Went in and added drivers to Left and Right channels until I had on each side for an 8 channel active speaker pair that could be defined broadly as a stand mount FAST 2.5 way with dual subs.
a. 1 tectonics BMR driver with a high pass (full range driver)
b. Two midbass drivers that had both a high and low pass that was the same. I had to disable the xover overlap check to achieve this. The dual midbass module is there for impact and additional SPL so in effect this is still to some degree a 2.5 way until you add subs.
c. 1 x M&K MX-70B. There are two, in opposite corners of my office. I have all bass under x hertz sent to them and each has a low pass.

3. Set up audiolense to take a measurement. WASAPI umik as my mic source, my sound card's asio driver for the speakers.
4. I let the channel assignments fall as they landed not knowing how they mapped. I executing 'check speaker connections' and mapped each channel number 0 through 7 to what driver the sound actually came out of. I did this at VERY LOW VOLUME so that I didn't damage the drivers and in this case since my tweeters are BMR's they could handle some pretty low frequencies, below 100hz so the risk of blowing them was very very low.
5. Changed all the channel numbers to reflect what reality was. I then took a few quick test sweeps at 3 seconds per driver while watching to make sure that the volume got me into the green range or a bit over at peaks but that no clipping occurred or signals with low boundaries. I noticed that some results were unsucessful or low. I ended up getting one after a lot of playing that resulted in 'EXCELLENT' on dynamic range and saved that measurement.
6. I chose to send it to my system.
7. I created a bunch of convolvers with the file format .alc at 96khz since that's what my system will operate at and only at 96khz. I tried a variety of corrections to test the gamut.
8. I loaded these into the audiolense convolver.

In doing this I confirmed that getting two channels to the convolver is all that you need and indeed the convolver itself contained the crossovers + corrections as well as the proper channel mappings that I defined in the measurement i.e. IT WORKS PERFECTLY.

I created some targets in linear and in minimum phase to try both. I looked online for examples and created and saved the following targets: Flat, Harman Curves in flavors of -4, -6, -8, and -10, B&K curve and some permutations of the Toole target variants.

Frequency procedure sounded good.
Minimum delay crossover sounded good and had some good dynamics.
Total Time Delay sounded so good that it felt strange. Adding Total Time Delay + per driver correction was my favorite. I found that minimum and linear phase seemed to have less audible impact to my ears than the differences between correction procedures did. In fact the per driver correction seemed to have more impact.

Remember, these are pretty new impressions for me spending a few in depth stints of tinkering so I'm an amateur myself but am pretty versed in the concepts having lurked and educated myself for a while.

You're preaching to the choir on getting this stuff to work together it's CRAZY painful.

I have DIRAC implemented on a Emotiva XMC-1 in another room as it has in-built Dirac Live capability. Dirac surface comparisons, and I'll do it from a pro/con perspective:

DIRAC:
PRO:
Almost idiot proof. You don't have to know anything beyond taking a measurement and drawing a curve.
Basically in my case it's a stand alone appliance.
I can apply it to my 2 channel stereo and also to my surround.
CON:
Zero control of what is happening in the magic box. You get what you get. The processing and algorithm are 'in the cloud'
Zero ability to create active speakers or crossovers
No export of anything but your target and maybe end result profile. You have to use a PC to switch what profile gets uploaded to the XMC1 and they have to live on the same network.
Software is immensely unstable.

AudioLense:
PRO:
The results are stellar. I don't know how they would compare to DIRAC, it would all depend on how well you executed your target and what correction procedure you choose.
You can customize your correction procedures.
You can save your profiles.
I can design xovers.
I can make totally almost idiot proof active speakers that crush my passive crossover compatriots with minimum knowledge - but not no knowledge.
Tons of flexibility in how you do this as long as you have a PC in the mix
I cannot express how long anything like this would take to even get a fraction of the results tinkering with REW, PEQ's in minidsp, etc. etc. I've tried. Not even close.
You can control the taps/precision of the filter which is awesome.
Con:
Requires some knowledge
Some mysteries still on how things work
May or may not have stability issues. I have had some oddness but I'm not sure if it's me or the software I'm reserving judgement. I clearly got great results but had a lot of oddities along the way.
Documentation, etc. has a long way to go. I got a lot of my learning from lurking on this forum.

Too much typing not enough listening. I'm going to go back to tinkering

 

[mccarty350]

BTW, I just noticed you're running the Andromeda II, I've looked at it several times on humble homemade hifi site it looks like a stellar project.

Also not sure if I mentioned my cheap office preamp that I got for under 200 shipped but the Sherbourn PT-7010 is a huge preamp/processor designed by Bob Carver and marketed rebadged as 3 different products. The Sherbourn PT-7010, the Emotiva DMC-1, and the Sunfire Theater Grand IV. All of them were 4k-7k in price at the time but are all useless for their purpose because they have no HDMI connections. However the SNR and fact that it has 8 analog channels of input via RCA and channels of RCA AND XLR out and that because the Sherbourn brand was mostly for professional installs and very little known you can get them off of ebay for easily under 300 shipped all day. They make a great volume control and 8 channel preamp with a lot of gain (I noticed you lose a TON of gain depending on your convolver, I've seen at least a 10db loss...at least...every time) that if you tried to compensate on your PC would often lead to clipping on the output signal but having that preamp with some heavy voltage gain (I think it kicked out 6 or 7 volts) will surely drive your amp to what it needs to get to. Just relating a recent option for my office system which is more of a living experiment and an okay place for non-top-of-the-line gear to live.

 

[Ofer]

Thanks for that. It is two systems, now I get it. So in your office you don't used the nVidia shield and route outside sources. Mapping in my case is no problem as 0 maps to 1 on the interface etc. Voltage is also not an issue as the RME outputs +4db in balanced and you can another extra 3db in the settings. I didn't understand this:
"6. I chose to send it to my system."
Send what to the system.
Also do you use virtual cable (not hifi cable) for system sounds and internet sound? or is it only voicemeeter for anything that isn't JRiver and foobar?
I am currently stuck. I will try to uninstall the ASIO bridge and reboot etc. But once I tried to add an ASIO strip to voicemeeter it seems that it circumvents the convolver all together. I get output on all 8 channels instead of 6 and they all play all the audible spectrum no xover. BTW I use the sample correction 3way file provided with AL as I haven't hooked the system yet .
The Andromeda was a great project and I was lucky to work with a genius carpenter. It was more then 10 years ago. Most of the time I used PA xover to drive it, Ashly. Lately I got RC mainly miniDSP. Your success inspires me to keep at it, as it feels like shooting in the dark. The option to go back to HTPC is there but it just isn't as family friendly as using a smart TV and google cast, there will be opposition.
Did you build the Linkwitz LX521? are those the speakers in your office?
Your subs look nice. Seems like good vfm.

 

[Ofer]

Another strange thing is when I press the 'Audiolense virtual sound card' in the voicemeeter settings it opens the RME settings.

 

[mccarty350]

That is bizarre, I swear, windows audio is full of such madness.

Yes, I have multiple systems and I'm always swapping stuff around and playing. IN my office I only am coming from a PC which makes life super easy.

Your RME is a beast and pricey, I'm not surprised that it has a ton of output.

#6 is my fault, what I mean to say was that I clicked the button that I think says 'send to current system' or some such that sends a completed measurement straight into your work space/displays the measurement as the current graph. Sorry about that.

I don't have to use anything to route my system sounds to Voicemeeter. It installed itself with the three virtual sound card drivers that I showed in an earlier post called like Aux, VAIO, and VAIO3. Set VAIO to be my default sound card and then boom all my system sounds go to it. You can also click on the task bar in the bottom right on your speaker icon and set it to point at VAIO. Then boom it all goes to voicemeeter that's all it takes. Then I manually configure jriver to point at AUX1 and manually configure Foobar to point at VAIO3. Honestly it doesn't matter, you could also just point all of them to VAIO as well, the only reason I point them to different virtual sound cards is that I can control their volume individually in Voicemeeter so that if I'm in a game I can make the music louder or software in comparison the game audio.

Unfortunately I don't know how the ASIO strips work as I've never really worked in professional audio software so I don't know it's impact or really what it does.

I was going to ask how the hell you did all the bracing, most of what it takes to make that speaker looks straight forward but the actual bracing looks complex. What I was always considering was the Nova Andromeda because I liked the drivers. Any way it goes it's a great candidate to just bypass the internal crossovers and actively amp it.

I'm pretty sure you can get this accomplished it's just going to take ridiculous tinkering. My success didn't come without tons of off irritation.

I did built the LX521's they are in my main room. Main room systems that I rotate in and out (that I need to sell down!):
0. My DIY wooden yuichi horn active 2 way that uses a Beyma neo compression driver (god I love wooden horns) and a 12" faitalpro bass bin and is ported. It's all modular so I can swap things in and out. Looking forward to running audiolense over this beast too!
1. Klipsch Chorus I
2. KEF 107 w/Kube
3. Polk SDA2B
4. Linkwitz LX521's
5. Klipsch CF-3
6. Bass subsystem for all is a pair of Dayton 18" reference series sealed subs powered by (like you) a pro amp. I have that pro amp hooked up to an alexa plug to verbally turn on and off as well as my two channel amp since I rarely use them and they have huge power switches.
Amps in that room: Ashley XLS402, McCormack DNA-225, pair of Outlaw 5000's for multichannel duty, and a big Sherbourn 12 channel amp for surround sound. Those big sherbourns go for next to nothing on ebay and are beasts, they also automatically turn off channels that have no input after 10 minutes which makes them cheap to run. Then I also have my XMC-1 which is the heart of everything, a minidsp 4x10, a minidsp HD, Behringer 8 channel audio interface (the sound card sounds better than this thing but it just plain works...I will likely sell it)

In my office...which is my tinkering room:
1. What i'm listening to now that is one of my DIY projects. It's a dayton 0.56 cubic foot enclosure. I put threated screw inserts where the baffle should go and then took the real baffles and used a straight router bit to make duplicates of it out of baltic birch. I then routed holes out for 2 x 6.5" woofers on one of the pairs of baffle plates and screwed them on. It's modular, I can take other baffles and route out new holes (or use a hole saw for tweeters) for other driver configurations. On the back of the baffle I have 2 x terminal cups that just go straight to whatever drivers I have. On top I have a PVC tube that has one end that's large and the other end that is smaller, it's a coupler. Between both holes I can accommodate all manners of driver sizes and cab the back with a piece of wood that I double sided tape or hot glue on (eitehr one comes off easy). This lets me have a totally modular DIY system. it eats up a total of 8 channels. Current drivers I am experimenting with are 3.5" tectonic BMR's, those nice VIFA TC9 full ranger drivers, a pair of morel MDT-29's, some cheap buyout foster ribbons, etc. So basically they day I got audiolense I dropped my crappy DIY jriver crossover and then let audiolense build an 8 channel xover based off of ballpark values and it came out STELLAR. I plan on swapping in and testing the other driver to see which I like.
2. IMF RSPM IV
3. Celestion SL600's
4. KEF 103.3
5. DIY: C-Note, Bagby Kairos, that HIvi 3.1a kit with the Sehlin mod, SB Acoustics Bromo diy kit, Mark Audio MS11 Alpair kit in 'bromo' boxes, ADS L710's, Usher Tiny Dancers (beryllium tweets), ADS 7le's (I think)....god knows what else. I have too much and I just need to sell. Audiolense with tectonic's and 2 x parts express woofers per box outperform them all. It sound stupid good.

I have too much stuff.

 

[Ofer]

Wow! you have and audio shop! Whenever I go to audio forums to try to find answers it always astounds me the level of equipment (and knowledge) people have. It is a good thing that there is no audio bar to access these forums otherwise I would surly be left out. BTW my RME is an interface that was discontinued in 2013. I got it on a deal for an odd ~400$. About the same price I paid for the 4x10HD, that was bought used.
Currently I have configured voicemeeter like you did and it looks like it sends stereo to all 8 channels.
There was a user named @aps that did exactly what I want to do using another RME unit. Unfortunately he is not responding. Guess once he got everything working he quite the forum.

 

[mccarty350]

I have an audio disease. Not uncommon. Nah, don't sell yourself short, you have obviously jumped into the deep end with both feet. You're building speakers, dealing with active speakers, working with best of breed audio products to make your own filters. This is the real frontier not just endlessly buying things.

If voicemeeter is sending it to 8 channels that risks some damage to active speakers and tweeters. I can't recall if I experienced the same behavior when I sent it straight to my sound card's output or not because I was too deep in it making changes. You've advanced into the bleeding edge without travelling through all the hoarding steps which is great.

Well...I hate to say this but some folks that you see that have stopped responding are no longer alive. I don't know if you ever hit audiokarma which is where the old timer vintage folks congregate but they lost so many foundational members in the last two years, some good friends to covid or old age. So sometimes when someone isn't responding it's because they are dead. I've experienced this a few times unfortunatley.

 

[Ofer]

Bleeding edge or not now nothing is working. Voicemeeter/AL convolver is not communicating with the audio interface at all. I won't be home for the coming week so it will rest. I might opt to saving some money and going the Mitch way. A lot of the hardcore well equipped computer audiophiles here when that route. It isn't cheap (for me) but it seems to bear result and he is well versed in all manners of computer audio. His HLC seems the way to go if you don't want to go with any special player.

It didn't occur to me the option of 'no longer with us'. The guy I was referring to was all about making the system family friendly so unlikely age is the issue. Hope it is not the case.