search

ebu / libear

A C++ library to render ADM content according to Recommendation ITU-R BS.2127
Apache License 2.0
27 stars 10 forks source link

[Question] Does libear support 3.1.2ch output? #51

Open sunghee-hwang opened 1 year ago

sunghee-hwang commented 1 year ago

For example, for a given input 7.1.4ch, can libear render it to 3.1.2ch? Where, 3.1.2ch is composed of L, C, R, LFE, Tfl, Tfr.

sunghee-hwang commented 1 year ago

I would like to request @tomjnixon to review this question.

tomjnixon commented 1 year ago

Probably not -- that's not a standard loudspeaker layout, and doesn't follow the conventions of the other loudspeaker layouts which would make it likely to work anyway.

The middle layer is incomplete, and the point source configuration procedure does not account for this.

If I needed such an output I'd render to 2+5+0 and use a downmix to map the M+-110 channels to the front.

sunghee-hwang commented 1 year ago

Thanks for the information and your recommendation.

felicialim commented 1 year ago

Thanks @tomjnixon for the quick response! Understood that this is not supported by the libear software.

Just wanted to check my understanding of your point about the point source configuration procedure not accounting for the incomplete middle layer.

Does this mean that ITU-R BS.2127 only supports rendering upper/middle/bottom layers which are complete, i.e. has a layout that is identical to one of the sound systems A - J?

tomjnixon commented 1 year ago

Does this mean that ITU-R BS.2127 only supports rendering upper/middle/bottom layers which are complete, i.e. has a layout that is identical to one of the sound systems A - J?

Officially, the supported output formats for BS.2127 are those listed in BS.2051 -- those are the layouts it was designed for, and has been validated with.

The point source component however does work for other layouts, as long as they are similar to the other ITU layouts in certain ways -- specifically they should have up to 3 layers of loudspeakers, and the convex hull should cover the origin of the coordinate system. The procedure that causes this is described in BS.2127 section 6.1.3, if you're interested.

felicialim commented 1 year ago

Thanks for the clarification.