EPS fails to import BS.2076-3 - files

[WernerBleisteiner]

audioFormatExtended version="ITU-R_BS.2076-3"> is not accepted in EPS 1.1.0
Manually fixed to audioFormatExtended version="ITU-R_BS.2076-2"> seems to
work - but creates only two empty audio tracks, instead of the contained 8 tracks
(2.0+5.1)
axml and screenshots attached.

adm-chna-axml-extracts.zip

But maybe the applied "Dolby E ADM and S-ADM Profile for emission" is to specific a use case to be considered for EPS anyway ;)

[firthm01]

Interesting... do you have an ADM WAV you could share? I've tried to handcraft one out of the AXML and CHNA you've given but I'm not sure if the issues I'm seeing are down to bad handcrafting or not.

[firthm01]

Hi Werner, On looking at the ADM and CHNA, something seems amiss here. All objects are using AudioPackFormat AP_00010002 (common definition stereo) but referring to AudioTrackFormats which have nothing to do with that AudioPackFormat in the CHNA. In fact none of those AudioTrackFormats exist at all except AT_00010012_01 which is for "PCM_TopSurroundRight" so not associated with the stereo layout.

That said, the EPS could really do with some better feedback on why it's dropping objects

[WernerBleisteiner]

Thanks for looking into this @firthm01.
I guess these ambiguities come from the fact, that the original encoded DolbyE (which is not contained here) folds 5.1 into a stereo stream.
ADM Usage Guidelines ITU-R BS.2388-5 #3.3.1 and #3.3.4 provide some hints how to handle encoded DolbyE.
However, this here should rather be treated like genuine 5.1.
We'll discuss that with the originators.

[dtatut]

@firthm01 & @WernerBleisteiner : the original file is old and broken. The purpose of the source file was to have a test file that fails on various QC/validation tests. It's probably worth to exchange some files to help the EPS development, what do you think?

[firthm01]

@dtatut Hi Dan, that probably makes sense but I'm not sure if there's the spare effort to contribute much. I'm not working at the BBC anymore so my so involvement is purely personal interest and I won't have much time to contribute myself. Probably worth a chat with Dave M or @rsjbailey.