Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mitchco]

Happy holidays!

Here is a stereo 4 way convolution cfg and filters for 192 kHz: https://accuratesound.blob.core.windows.net/sotadrc/stereo_4way_convolution.zip

The routing of the digital XO is:
Channel 1 and 2: bass
Channel 3 and 4: lower mids
Channel 5 and 6: upper mids
Channel 7 and 8: treble

JRiver convolution works correctly. However, loading Hang Loose Convolver (HLC) as a VST3 convolution plugin, channels 7 and 8 (treble) are swapped with channels 5 and 6 (upper mids). Note that HLC output meters are showing the correct channel routing/output, but after convolution, the channels are swapped.

Plugging HLC into another Host application, the channel routing/output is correct.

The way the output format is set is attached.

Thanks for any assistance.

[Matt]

If you pick 7.1 instead of extra channels does it make a difference?

[Mitchco]

Hi Matt,

7.1 produces the same swapped channels. Thanks.

[Matt]

Thanks.

We're not doing any channel swapping that I can see.

I asked the author of HLC for help and maybe they'll know more.

Thanks again.

[whoareyou]

I think this has something to do with JUCE.

With a convolution I used for a 5.2 system, I configured output configuration 7.1 (which I believe is equivalent to 2 with 6 extra).  For this system the channel order went from L,R,C,SW,SL,SR to L,R,C,SW,RL, RR, SL,SR for the 7.1 channel mapping.  So, in JRiver I swapped the RL, RR, SL, SR channels to make this work with the 5.1 input material to match the 7.1 channel order

I think I originally performed this swap after the convolution. 

Now, this is the interesting part.... At some point the JUCE library was updated.  After the JUCE update the channel order swap no longer worked correctly (nothing changed in JRiver) and to fix it, I believe I had to move the channel swap to before the convolution.  It seemed as though the JUCE library was changing the channel order. 

(Don't hold me to where the JRiver channel swap actually took place.  I just know that the channel order behaved differently depending on the version of JUCE).   My Dirac VST3 DSP chain also had to be modified after they updated JUCE versions for their VST.

Possible solution is to try the 5.1.4 mapping ..... When I started using my 10 channel setup for Atmos, I switched to the 5.1.4 mapping.  For whatever reason, this output configuration requires no channel swaps.   So, I would try the 5.1.4 mapping and see if that can get you going. 

Seems a little weird but ....

[Matt]

With Mitch's help we are going to switch how 8 and 9 channel audio is reported to VST3 plugins. This should solve this problem. Feedback appreciated. Thanks.

[HamishNorton]

My thanks to Mitch and Matt for fixing this issue at the interface between two software packages over Christmas break faster than I would have thought possible. Amazing.

[Matt]

Build public:
https://yabb.jriver.com/interact/index.php/topic,140330.0.html

Please let us know how you fare. Thanks.