Dynamic range compression(DCR) [volume normalization] on MCU

[khaled]

we are developing a webRTC stack based on Janus as MCU/SFU and coturn as TURN server. The media stream is a one-way AV stream, sourced from mobile-device and the subscriber(viewer) is on browser. The whole MCU + TURN + HAproxy stack is on AWS. Given the stack, our avg. latency from publisher to subscriber is ~400ms as the distance is half-the-globe, for now which is okay. 

Now, the issue we are facing is that on publish end, usually two persons are speaking with quite a bit of volume difference in their voice making it hard to comprehend for the listener. So, we are thinking how we can apply Dynamic range compression aka volume normalization on the stream.

Now, what would be the best approach to solve it? Is it to do it through a server-side filter-plugin on the Janus-MCU? what'd be the impact on the latency? Or, it is wise to do it on the client-side on browser through customizing webRTC playback.

any pointer or hints would be greatly helpful.

// khaled

[khaled]

i did forget to mention the codec part. 

We are currently using VP8 & PCMU codec for video and audio encoding respectively.

// khaled

[Lorenzo Miniero]

What do you mean by SFU/MCU? Which plugins are you using?

L.

[khaled]

We are using the Videoroom plugin. 

Can we use an DCR audio filter with it? what can be the impact on overall latency?

// khaled

On Monday, November 14, 2016 at 5:32:26 PM UTC+6, khaled wrote:

[Lorenzo Miniero]

The VideoRoom plugin doesn't process media, it just forwards it exactly as it is. If you want to do processing, you'll have to either do that on the client side (e.g. in the browser, if possible) or modify the plugin to process the audio somewhere in there.

Lorenzo

[khaled]

Thanks a lot for the input, Lorenzo.

we are investigating both doing it on browser and on server-side. 

From your experience, can you give a guesstimate (even a wild one will help) that how much latency it can add if it is done on the server-side?

// khaled

[Lorenzo Miniero]

No idea, sorry, not familiar with the approach you proposed.

L.

--
You received this message because you are subscribed to the Google Groups "meetecho-janus" group.
To unsubscribe from this group and stop receiving emails from it, send an email to meetecho-janus+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.