Focusrite ASIO/Extreme low latency Support

[Timothy4Real]

Hello, may your days be well!

 

I'm having quite the awful time configuring ASIO for my focusrite 8i6, which I can use spdif as an output, but it seems to crash after applying, no matter what settings I use. It's awful inconvenient.

 

If someone could help me utilize all of my inputs with WASAPI, that would help me greatly as well, especially as I have read very little positive about ASIO support. 

 

My use case is that me and my friend wish to play music across the country with as little of delay as possible, and we have paid for high bandwidth, low-ping internet connections to do this plus video, and we have built the servers to be running on multiple machines so as to reduce processing times. 

 

I have reduced the latency to a very small amount, but it does affect timing of instruments. 

 

Please know that we understand that noise is not going to be filtered out and this is not for studio level recording. I have found a good balance of quality versus latency at 44kb/s, and have gone as far down as 8kb/s and had the same level of delay - in loopback tests.  

 

This may be an engineering/physics level problem, and may not actually ever work, but I know there is a community of musicians who would benefit from extreme low latency audio sessions and would like my voice to be heard for that reason alone. 

 

ASIO may be the wrong direction all together, and I'm fine being pointed in the right direction, so please excuse my lack of technical knowledge and know that I have taken a scientific level of testing to find the best options in Mumble as they're presented to me. 

 

Thank you for any responses. 

[Krzmbrzl]

As far as I know the ASIO implementation in Mumble is quite buggy and unstable, so I would not recommend it. See e.g. https://github.com/mumble-voip/mumble/issues/2274, which seems kinda similar to what you are describing.

 

What is the issue you are seeing with WASAPI? What do you mean by "all your inputs"? Mumble only processes one input stream, so if you have multiple streams that you want to send, you'll have to use a mixer (software or hardware) before feeding the audio input into Mumble.