Centova Technologies Inc.

Products and Services

Centova Cast
Customer Support
Centova Technologies
 

Centova Technologies Forum

Author Topic: FLAC, AAC, and OGG support in autodj  (Read 882 times)

DJ Nativus

  • Newbie
  • *
  • Posts: 5
    • Furry FM
FLAC, AAC, and OGG support in autodj
« on: November 04, 2016, 08:25:31 pm »
So certain engine software for AutoDJ, such as LiquidSoap, comes with AAC, OGG, and FLAC support but Centova only recognizes MP3. Sad! Competitors like SAM Broadcast Cloud support this.

Why would you want formats like FLAC in autodj? Because when re-encoding tracks, it maintains much better quality and if you move your station from say 128 kbps to 192 kbps, you don't have to reupload all the tracks. (And it's sinful to transcode to a higher bitrate than the original file).

OGG and FLAC use similar tagging system and Liquidsoap already supports the formats, so should be pretty easy to implement.

Would love to see that!

Mr_NightRyder

  • Newbie
  • *
  • Posts: 26
Re: FLAC, AAC, and OGG support in autodj
« Reply #1 on: May 03, 2017, 09:31:03 pm »
Consider this upvoted I can do it with my whmsonic it needs to be capable in centova also.

aleph

  • Newbie
  • *
  • Posts: 13
Re: FLAC, AAC, and OGG support in autodj
« Reply #2 on: August 23, 2017, 07:57:44 pm »
In my testing, I found that the fairly unknown Opus format produces even better results than AAC+. Since Centova can broadcast - and therefore encode - in AAC and Opus, why not implement the easier part and implement a similar decoding option?

Roger

  • Centova Staff
  • *
  • Posts: 356
Re: FLAC, AAC, and OGG support in autodj
« Reply #3 on: August 25, 2017, 09:19:02 am »
Hello DJ Nativus,

Regarding your comments:

So certain engine software for AutoDJ, such as LiquidSoap, comes with AAC, OGG, and FLAC support but Centova only recognizes MP3. Sad!

The main obstacle for this feature it's not that of being able to decode different compression formats (as that would be relatively easy to do), but handling the non-standardized metadata headers in OGG containers -- I don't even know what is supported by FLAC in terms of metadata, however FLAC presents also the issue of CPU usage which is likely to upset streaming providers to the point that they will likely just turn it off themselves --. AAC and AAC+ do support ID3, however most existing AAC media files are encoded using VBR (variable bit-rate) which is known to cause problems with Shoutcast/Icecast streaming.

Bare in mind that ID3 metadata, is essentially the basis for Centova Cast's media library and statistics features, and that supporting a second form of metadata (specially a non-standardized one) would require a huge rework of this two features.

Competitors like SAM Broadcast Cloud support this.

SAM is not a competitor of Centova Cast at all, one is a desktop application while the other one is a server side administration platform.

Why would you want formats like FLAC in autodj? Because when re-encoding tracks, it maintains much better quality and if you move your station from say 128 kbps to 192 kbps, you don't have to reupload all the tracks. (And it's sinful to transcode to a higher bitrate than the original file).

Actually, you can already do this by using high bit-rate mp3 files, for example 320kbps (in fact we often recommend this when clients complain of low audio quality in their streams).

The difference between down-sampling from a FLAC or a 320kbps mp3 file to a 128kbps mp3 stream is negligible.

OGG and FLAC use similar tagging system and Liquidsoap already supports the formats, so should be pretty easy to implement.

Would love to see that!

Again, while this is true for the decoding part, the post processing of metadata and accuracy of features like royalty reports will be quite difficult to match considering the non-standardized metadata of this media containers. Maybe if this formats become hugely popular in the future (which currently they are not except for AAC), and the metadata becomes fairly standardized.



Regards.
« Last Edit: August 25, 2017, 09:29:29 am by Roger »