[Feature request] Downmix using HRTF (aka. virtual surround for headphones)

[Sonicity]

Hi,

I enjoy my purchase and use of Voicemeeter. The downmix feature is useful when I want to preview materials composed in 7.1 format.
However I wish we could use HRTF to downmix 7.1 to stereo (aka. virtual surround for headphones).

The principle of HRTF is that, our ear canal is a filter that has different responses for sound coming from different azimuth and altitude. Also the signal coming from different distances reaches both ears with different latency.
We apply HRTF to the audio stream to simulate an ear canal, allowing our brain to feel "sound coming behind us" through any stereo headphones.

It is not difficult to include HRTF into Voicemeeter because we already have an EQ, and HRTF is just another EQ + Delayer.
Third-party libraries exists: FFmpeg (libavfilter.dll) already contains an LGPL implementation named "SOFAlizer" which can be directly included into Voicemeeter if license permits. Sofacoustics also provides an LGPL-licensed SOFAlizer implementation in C.

Although HRTF data is personalized for ear shape, we can ignore that in most cases. It might be possible if the user can import their personalized HRTF data in SOFA format, but an average HRTF data is enough to for music and games.

I am looking forward to your reply, and I am willing to help with problems regarding developing this feature.

Thanks!

[Sonicity]

I apologize for bumping this old thread, for the fear that it may be ignored if dropped to Page 2.

I understand it's hard to add new features, but I would like to request a comment on this topic.
Will it be rejected? Or accepted but delayed after other prioritized things?

Thank you.

[Vincent Burel]

Voicemeeter provides and API to make APA (Audio Processing as Application) . we already use it to make a 15 bands graphic EQ or 8x8 gain matrix to process a selected BUS in Voicemeeter (installed with Voicemeeter).

We can invite guys of SOFALIZER or SOFACOUSTIC to make their application for Voicemeeter base on our SDK or sample example given there: viewtopic.php?f=8&t=394

[Sonicity]

Thank you for your reply!

I currently use a commercial HRTF software (Razer Surround) and chain it through Voicemeeter to achieve the same result (but with huge latency).

I have briefly read the Voicemeeter API documentation PDF and understood that there are 3 places where I can insert filters. I thought I can write an adapter from LADSPA API to Voicemeeter APA, therefore I can use the existing LADSPA plugin for HRTF in Voicemeeter.

May I ask if anywhere already exists any adapters to insert a standardized plugin system to the 3 places? For example:

• VST (widest support on Windows)
• LADSPA (featuring extremely simplistic C API, both host side and plugin side, origin from Linux but ported to Windows so we can use it)
• JACK (featuring computing graph, predictable latency with scheduler-level realtime optimization)

With the help of API adapters, it will be very easy for me to chain them together and write an HRTF downmix APA for Voicemeeter.

Additionally, even if we ask people from SOFACOUSTIC to write a plugin for Voicemeeter, they might would still expect an industrial standardized API such as LADSPA or VST.

I really hope we can include built-in "HRTF Downmix" along with "Downmix A" and "Downmix B" to showcase the 7.1 processing abilities of Voicemeeter, because usually it has higher fidelity to use an HRTF-enabled headphone to make music than to arrange 8 speakers in a small room at inaccurate angles.

[Vincent Burel]

Voicemeeter API is very easy to implement in a regular user application (can be done in one day).
i don't know if it will become a standard, but i know that there is already millions of users, and more again every year...

[Sonicity]

That's cool!

Since the API is easy, I will try to proceed with this feature on my own during holidays.

Thank you for your detailed answers!