Re: [FFmpeg-devel] Fixes for audio/L16

2018-11-24 Thread Carl Eugen Hoyos
2018-11-24 10:27 GMT+01:00, Igor Derzhavin : > Do you accept these patches? Or what can I do more? > We use ffmpeg (libavformat/libavcodec) within our company and came across > that http url with content type > "audio/l16;rate=24000;channels=1;endianness=big-endian" incorrectly opens > as s16le wi

Re: [FFmpeg-devel] Fixes for audio/L16

2018-11-24 Thread Paul B Mahol
On 11/22/18, Igor Derzhavin wrote: > Hello! > A little bunch of fixes for audio/L16 mime type handling: > 0001 - RFC 2045 says MIME type should be case insensitive; > 0002 - RFC 2586 says audio/L16 should be in network byte order (aka big > endian); > 0003 - though "endiannes" parameter not in RFC

Re: [FFmpeg-devel] Fixes for audio/L16

2018-11-24 Thread Igor Derzhavin
Hello! Do you accept these patches? Or what can I do more? We use ffmpeg (libavformat/libavcodec) within our company and came across that http url with content type "audio/l16;rate=24000;channels=1;endianness=big-endian" incorrectly opens as s16le with sample rate 16000. This patches fixes that. O

Re: [FFmpeg-devel] Fixes for audio/L16

2018-11-22 Thread Igor Derzhavin
No, I don't now applications that can play pcm streams through HTTP. VLC misdetect such streams as mp3, and there is a two years old bug https://trac.videolan.org/vlc/ticket/17229. On Thu, Nov 22, 2018 at 4:00 PM Carl Eugen Hoyos wrote: > 2018-11-22 10:16 GMT+01:00, Igor Derzhavin : > > > A litt

Re: [FFmpeg-devel] Fixes for audio/L16

2018-11-22 Thread Carl Eugen Hoyos
2018-11-22 10:16 GMT+01:00, Igor Derzhavin : > A little bunch of fixes for audio/L16 mime type handling: > 0001 - RFC 2045 says MIME type should be case insensitive; > 0002 - RFC 2586 says audio/L16 should be in network byte order (aka big > endian); > 0003 - though "endiannes" parameter not in RF

[FFmpeg-devel] Fixes for audio/L16

2018-11-22 Thread Igor Derzhavin
Hello! A little bunch of fixes for audio/L16 mime type handling: 0001 - RFC 2045 says MIME type should be case insensitive; 0002 - RFC 2586 says audio/L16 should be in network byte order (aka big endian); 0003 - though "endiannes" parameter not in RFC it is widely used; From e40e9567995042d75f8070d